Вариация структурных и оптических свойств оптически чувствительных нано- и микроразмерных кристаллов с помощью микрофлюидных технологий тема диссертации и автореферата по ВАК РФ 00.00.00, кандидат наук Корякина Ирина Георгиевна

• Корякина Ирина Георгиевна
• кандидат наук
• 2023

Оглавление диссертации кандидат наук Корякина Ирина Георгиевна

Реферат

Synopsis

Introduction

CHAPTER 1. Optically responsive nanocrystals and methods of its synthesis

1.1. Optically responsive crystals

1.1.1. Perovskites

1.1.2. Conventional synthesis methods of perovskites

1.1.3. Metal-organic frameworks

1.1.4. Conventional synthesis methods of MOFs (tunability of intrinsic properties)

1.2. Formation mechanisms of nanoparticles

1.3. Overview of Microfluidics

1.3.1. Theory of microfluidics

1.3.2. Materials of microfluidics

1.4. Microfluidics for materials synthesis

1.4.1. Perovskite synthesis via microfluidics

1.4.2. MOF synthesis using microfluidics

Conclusion on chapter

CHAPTER 2. Fabrication and characterization

2.1. Materials

2.1.1. Materials for perovskite synthesis

2.1.2. Materials for MOF synthesis

2.3 Microfluidic chip fabrication

2.3.1. Soft lithography

2.4. Synthesis setup

2.4.1. Continuous flow

2.4.2. Droplet microfluidics

2.5. Structural characterization technologies

2.5.1. Electron microscopy

2.5.2. Powder X-ray diffraction (pXRD)

2.6. Optical characterization technologies

2.6.1. Photoluminescence (PL)

2.6.2. Quantum yield of photoluminescence (QYPL)

2.6.3. Lasing

2.6.4. Raman-spectroscopy

2.7. Dye loading

2.8. Cells study

2.8.1. Cells dye loading

2.8.2 Toxicity study

2.8.3. Uptake study

Conclusion on chapter

CHAPTER 3. Continuous flow synthesis

3.1 Continuous flow synthesis of MOF

3.1.1. Product yield estimation

3.1.2. Mixing in the channel of MFC

3.1.3. Structural characterization of MOF HKUST-1

3.1.4. HKUST-1 defects study

3.1.5. Defects of HKUST-1 for dye adsorption

3.2 Continuous flow synthesis of perovskites

3.2.1. Structural characterization of CsPbBr3 perovskite particles

3.2.2. Optical characterization of CsPbBr3 perovskite particles

Conclusion on chapter

CHAPTER 4. Droplet microfluidics synthesis

4.1 Droplet microfluidics synthesis of perovskites

4.1.1. Temperature influence on CsPbBn perovskite structural properties

4.1.2. Hydrophilic-lipophilic balance influence on CsPbBr3 perovskite structural properties

4.1.3. Optical properties of the obtained CsPbBr3 perovskites

Conclusion on chapter

Conclusion

List of abbreviations

Bibliography

Acknowledgments

Appendix A

Реферат

Общая характеристика диссертации

Введение диссертации (часть автореферата) на тему «Вариация структурных и оптических свойств оптически чувствительных нано- и микроразмерных кристаллов с помощью микрофлюидных технологий»

Актуальность темы

Оптически чувствительные материалы представляют собой класс материалов, проявляющих уникальные свойства под воздействием излучения. В зависимости от своей природы такие материалы обладают различными оптическими функциональными возможностями. Ярким представителем класса оптически чувствительных материалов являются металл-органические каркасы (МОК). Они представляют собой органико-неорганические гибридные кристаллические пористые материалы, которые имеют большую площадь внутренней поверхности. МОК обладают однородной структурой пор, регулируемой пористостью и химической функциональностью. Эти параметры позволяют успешно варьировать структуру каркаса и его функциональность, что, в свою очередь, позволяет создавать материалы, которые избирательно поглощают молекулы газа [109; 127], красителя [99] или лекарства [81; 88; 134] при воздействии на МОК оптическим излучением определенной длины волны. Наиболее распространенными методами синтеза МОК являются химические методы - сольвотермальный и микроволновый.

Другими оптически чувствительными материалами, представляющим интерес для применения в фотонике и оптоэлектронике, являются галогенидные перовскиты CsPbXз (X = С1, Вг и I). Наноструктурированные перовскитные материалы вызывают интерес благодаря таким свойствам, как перестраиваемая ширина запрещенной зоны, возможность излучения во всем видимом спектре, узкая ширина линии люминесценции, а также многофотонное поглощение. Кроме того, перовскиты из галогенида свинца различной формы (сферы, кубы, нитевидные структуры, диски, кольца и т.д.) могут служить резонансными полостями с активной оптической средой для разработки микромасштабных фотонных устройств [71]. Традиционные подходы к изготовлению CsPbXз (X = С1, Вг и I) включают лазерную литографию [90], химическое осаждение из паровой

фазы в инертной атмосфере [31], а также высокотемпературную инжекцию, что времязатратно, чувствительно к условиям окружающей среды и требует дорогостоящего оборудования.

Вышеперечисленные методы синтеза как перовскитов, так и МОК оперируют макропараметрами, что отражается на качестве синтезируемых материалов. Следует подчеркнуть, что физико-химические свойства материалов (фаза, размеры, морфология и состав) в основном определяются процедурой их синтеза, которая обычно включает в себя несколько этапов при использовании традиционных методов синтеза. Эти методы не всегда воспроизводимы из-за недостаточного контроля тепломассопереноса.

Для преодоления перечисленных недостатков предлагается использовать микрофлюидные технологии для синтеза оптически чувствительных материалов (перовскитов, МОК).

Микрофлюидные технологии оперируют жидкостями микро- и нанообъемов, что позволяет повысить точность реакции, варьируя такие параметры как концентрация реагентов и время реакции. Высокая скорость тепломассообмена в каналах микрофлюидного чипа значительно снижает время проведения синтеза кристаллов. А благодаря малому масштабу в микрофлюидике проще получить перенасыщение, чем при кристаллизации в традиционных методах, и, следовательно, возможно достигнуть кристаллов меньшего размера за счет увеличения скорости образования зародышей. Также, микрофлюидика позволяет создать условия для более гомогенной среды, что способствует равномерному распределению зародышей и более равномерному росту кристаллитов [113].

В микрофлюидике различают непрерывный тип потока жидкостей, сегментированный поток, а также деление потока на капли. В случае непрерывного типа потока смешивание жидкостей происходит за счет диффузии на границе между ними. А в случае сегментированного потока и потока, разделенного на капли - жидкость разделена на объемы ограниченного объема, то есть на так называемые микрореакторы. В случае капельной микрофлюидики дисперсная фаза окружена

сплошной фазой, в то время как в случае сегментированного потока дисперсная фаза взаимодействует со стенкой микрофлюидного устройства.

В работах [138] и [24] рассматривается применение микрофлюидных технологий для синтеза материалов. Так, применение метода капельной микрофлюидики позволяет точно регулировать время реакции, что отражается на размере синтезируемых МОК MOF-5 [23]. В работе [115] Чжан и др. провели синтез нитевидных наноразмерных CsPbBгз в микрофлюидном устройстве непрерывного потока. Для этого прекурсоры перовскита смешивали контролируемым образом, и время реакции регулировалось скоростью потока. Морфология полученных перовскитов и их оптические характеристики зависели от скоростей потока и температуры реакции. Авторы демонстрируют, что разработанный микрофлюидный подход способствовал выравниванию удлиненного CsPbBгз, в то время как обычный синтез методами «мокрой» химии приводит к получению нитевидных структур неровной формы при тех же условиях.

Таким образом, микрофлюидные технологии позволяют преодолеть недостатки традиционных методов синтеза материалов и точнее регулировать параметры кристаллизации (кинетику кристаллизации, скорость кристаллизации, и т.д.) и условия формирования физико-химических свойств материала, что позволяет контролировать структурные и оптические свойства материалов.

Диссертационная работа посвящена экспериментальному исследованию зависимости структурных и оптических свойств оптически чувствительных материалов (перовскит, МОК) от условий синтеза при применении микрофлюидного метода синтеза.

Основной целью диссертации является разработка новых методов контролируемого синтеза оптически чувствительных нано- и микроразмерных кристаллов (перовскитов, МОК) путем применения микрофлюидных методов синтеза для получения материалов с определенными оптофизическими свойствами, а также изучение структурных и оптических свойств полученных структур.

Для достижения данной цели в рамках диссертации были поставлены и решены следующие задачи:

Задача 1. Экспериментальная демонстрация влияния температуры синтеза при применении микрофлюидного метода синтеза на структурные свойства (дефектность, линейный размер, размер кристаллита) МОК;

Задача 2. Экспериментальная демонстрация влияния параметров микрофлюидного синтеза перовскитов CsPbBrз в непрерывном потоке на структурные (линейный размер, морфология, кристаллическая фаза), а также оптические (генерация вынужденного излучения) свойства кристалла.

Задача 3. Исследование структуры (линейный размер, морфология, кристаллическая фаза) перовскита CsPbBrз при вариации величины гидрофильно-липофильного баланса при синтезе методом капельной микрофлюидики.

Основные положения, выносимые на защиту:

1. Металлоорганические каркасы НКШТ-1 (С18ШСЩО12), полученные методом микрофлюидного синтеза в непрерывном ламинарном потоке при смешении органического лиганда и соли металла с расходом жидкости равном 3-3.5 мкл/мин, в диапазоне температур от 0 до 110 град., обладают точечными и поверхностными дефектами, стимулирующими усиленную в 2 раза сорбцию органических красителей родамина Б и цианина-5 в сравнении с микрокристалами НКЦБТ-1, полученными сольвотермальным синтезом.

2. Формирование кристаллов CsPbBrз определяется соотношением смешиваемых в ламинарном потоке 2-пропанол-вода и раствора прекурсоров перовскита и в соотношении 1:1.1 приводит к образованию нитевидных кристаллов с высоким структурным совершенством и низкой концентрацией дефектов кристаллической решетки, что подтверждается

квантовым выходом фотолюминесценции до QYPL = низкопороговой лазерной генерацией равной 15.7 мкДж/см2.

90.3% и

3. Кинетика роста кристаллов CsPbBn определяется объемом микроразмерного реактора, сформированного посредством контролируемой эмульсификации раствора прекурсоров перовскита в несущей среде внутри канала микрофлюидного чипа, что позволяет получать перовскитные монокристаллы кубоидной формы со средними размерами из диапазона от 120 нм до 650 нм, которые демонстрируют квантовый выход фотолюминесценции до 86%.

Научная новизна работы отражена в следующих пунктах:

1. Впервые продемонстрирована повышенная поверхностная дефектность металл-органических каркасов HKUST-1, полученная при применении микрофлюидного синтеза в непрерывном потоке и проявляющаяся в увеличении адсорбции молекул красителя.

2. Впервые получены микро- и нано- структуры перовскита кубической и нитевидной формы в суспензии, что позволяет помещать микро- и наночастицы на структурированные подложки для улучшения оптических свойств.

3. Впервые продемонстрирована вариация размера структур перовскита CsPbBr3 при изменении гидрофильно-липофильного баланса поверхностно-активных веществ при синтезе посредством метода капельной микрофлюидики.

Практическая значимость результатов диссертационной" работы состоит в разработке технологии синтеза оптически чувствительных микро- и наноструктур микрофлюидным методом. Разработанный метод синтеза позволяет получать

оптически чувствительные кристаллы с заданными структурными характеристиками (морфологией и размером) и оптическими характеристиками. Такой подход к синтезу обеспечит одновременно высокую монодисперсность и воспроизводимость синтеза оптически чувствительных кристаллов.

Достоверность полученных результатов обеспечивается широкой вариацией используемых в работе современных методов исследования, и подтверждена многократной воспроизводимостью экспериментальных результатов разработанных методов, а также воспроизводимостью измерений структурных и оптических характеристик с изучаемых материалов. Результаты, полученные в ходе выполнения диссертации не противоречат ранее опубликованным работам, апробированы на международных научных конференциях, а также опубликованы в международных рецензируемых журналах.

Апробация работы.

Основные результаты работы докладывались и обсуждались на следующих конференциях:

• Lab-on-a-Chip and Microfluidics Asia 2022, Токио, 6-7 октября 2022

• XI конгресс молодых ученых ИТМО, Санкт-Петербург, 4-8 апреля 2022

• VI International Conference on Metamaterials and Nanophotonics, Metanano 2021, онлайн, 13 - 17 сентября 2021

• School on Advanced Light-Emitting and Optical Materials (SLALOM), онлайн, 28-30 июня 2021

• Сессия-форум «Компьютерный инжиниринг в трансформации традиционных индустрий 2020», онлайн, 17 декабря 2020

• IX конгресс молодых ученых ИТМО, Санкт-Петербург, 15-18 апреля 2020

• V International Conference on Metamaterials and Nanophotonics, Metanano 2020, онлайн, 14 - 18 сентября 2020

• IV International Conference on Metamaterials and Nanophotonics, Metanano 2019, Санкт-Петербург, 15 - 19 июля 2019

Личный вклад автора.

Личный вклад автора состоит в изготовлении микрофлюидных устройств, проведении экспериментов по синтезу перовскитов и металл-органических каркасов, обработке экспериментальных данных по результатам рентгеноструктурного анализа, а также по результатам электронной микроскопии (сканирующей, просвечивающей). Автор принимала участие в измерении спектров комбинационного рассеяния, в проведении измерений оптических параметров (лавирование с синтезированных структур). Помимо этого, автор принимала активное участие в подготовке текста научных статей и лично представляла результаты работы в виде докладов на научных конференциях.

Структура и объем диссертации.

Диссертация состоит из введения, четырех глав и заключения. Полный объем диссертационной работы составляет 211 страниц, включая библиографический список из 138 наименований. Работа содержит 43 рисунка и 11 таблиц.

Публикации.

Основные результаты по теме диссертации изложены в 8 публикациях. Из них 8 изданы в журналах, индексируемых в базе цитирования Scopus.

В международных изданиях, индексируемых в базе данных Scopus:

1. I. Koryakina, S. Bikmetova, D. Khmelevskaya, D. Markina, A. Kuleshova, L. Logunov , A.Timin, A. Pushkarev, S. Makarov, Mikhail V. Zyuzin. Droplet Microfluidic Synthesis of Halide Perovskites Affords Upconversion Lasing in Mie-Resonant Cuboids //ACS Applied Nano Materials. - 2023. - Т. 6. - №№. 6. - С. 43704378.

2. Koryakina I., Bachinin S., Gerasimova E., Timofeeva M., Shipilovskikh S., Bukatin A., Timin A., Milichko V. & Zyuzin M.V. Microfluidic synthesis of metal-

organic framework crystals with surface defects for enhanced molecular loading //Chemical Engineering Journal. - 2023. - Т. 452. - С. 139450.

3. Koryakina I., Afonicheva P., Arabuli K., Timin A., Evstrapov A. & Zyuzin M. V. Microfluidic synthesis of optically responsive materials for nano-and biophotonics //Advances in Colloid and Interface Science. - 2021. - Т. 298. - С. 102548.

4. Koryakina I., Bikmetova S., Arabuli K., Evstrapov A., Pushkarev A., Makarov S. & Zyuzin M. V. Continuous-Flow Synthesis of Perovskite Particles for Optical Application //Journal of Physics: Conference Series. - IOP Publishing, 2021. - Т. 2015. - №. 1. - С. 012072.

5. Koryakina I. G., Naumochkin M., Markina D. I., Khubezhov S. A., Pushkarev A. P., Evstrapov A. A. & Zyuzin M. V. Single-Step Microfluidic Synthesis of Halide Perovskite Nanolasers in Suspension //Chemistry of Materials. - 2021. - Т. 33. - №2. 8. - С. 2777-2784.

6. Koryakina, I., Markina, D., Evstrapov, A., Pushkarev, A., Makarov, S., & Zyuzin, M. V. Microfluidics-based synthesis of lead cesium bromide perovskite microcrystals //AIP Conference Proceedings. - AIP Publishing, 2020. - Т. 2300. -№. 1.

7. Koryakina, I., Pushkarev, A., Makarov, S. V., Zyuzin, M. V., & Evstrapov, A. A. Synthesis of perovskite nanoparticles in microfluidic chips //Journal of Physics: Conference Series. - IOP Publishing, 2020. - Т. 1461. - №. 1. - С. 012071.

8. Koryakina, I., Kuznetsova, D. S., Zuev, D. A., Milichko, V. A., Timin, A. S., &

Zyuzin, M. V. Optically responsive delivery platforms: From the design considerations to biomedical applications //Nanophotonics. - 2020. - Т. 9. - №. 1. - С. 39-74.

ОСНОВНОЕ СОДЕРЖАНИЕ РАБОТЫ

Во введении диссертации сформулированы цель, задачи и основные выносимые на защиту положения. Обоснованы актуальность, достоверность и значимость работы, а также описаны новизна и личный вклад автора.

В первой главе представлен обзор литературных источников, посвященный традиционным методам синтеза таких оптически чувствительных материалов, как металл-органические каркасы и перовскиты, а также зависимости структурных и оптических свойств от условий синтеза [А1, А2]. В главе также рассмотрены текущие разработки по улучшению процедуры синтеза для получения разнообразных свойств МОК и перовскитов, что подтверждает актуальность рассматриваемой работы и правильность сформулированных задач.

Во второй главе приведены экспериментальные методики получения и исследования образцов, используемые в работе. Описаны материалы, использованные в работе, а также подробно приведена характеризация и параметры методов анализа образца, которые были использованы в работе. Также, глава включает использованные геометрии микрофлюидных чипов и схему установок синтеза.

Третья глава диссертационной работы посвящена экспериментальному изучению синтеза оптически чувствительных нано- и микрочастиц в непрерывном потоке микрофлюидного чипа.

Первая часть третьей главы посвящена синтезу металл-органических каркасов (МОК). Присущие МОК свойства, такие как пористость и химическое разнообразие, позволяют использовать их в качестве систем доставки лекарств. При этом развитая поверхность МОК позволяет исследовать новую форму эффективной доставки с дистанционным управлением. В разделе §3.2 описан синтез микрокристаллов МОК (НК^Т-1) с развитыми поверхностными дефектами в непрерывном потоке МФЧ (Рисунок 1) [А3].

Рисунок 1 - Схема микрофлюидного синтеза МОК и его дальнейшего использования для доставки лекарственных средств. Этапы исследования: (i) микрофлюидный синтез МОК, (ii) структурная и оптическая характеристика, (iii) загрузка красителя и его высвобождение при лазерном облучении и (iv) поглощение клетками и тесты на токсичность in vitro.

Проведена вариация синтеза для получения желаемого МОК, который обладает необходимыми параметрами для повышенной загрузки молекул за счет повышенной дефектности поверхности МОК. В качестве параметра вариации была выбрана температура. Таким образом, проведен синтез МОК при температурах 0-22-50-70-90-110оС. Проведены структурный анализ (методами СЭМ, РСА) и Раман-спектроскопия (Рисунок 2), которые указали на наличие поверхностных дефектов.

О 250 500 750 1000 1250 1500 1750

Raman shift (cm"1)

pm цгп

Рисунок 2 - Оптическая характеристика полученных МОК. (А) Раман-спектроскопия НКи8Т-1 (вставка: оптическое изображение монокристалла НКи8Т -1); (Б) картирование с шагом 100 мкм спектров комбинационного рассеяния света монокристалла НК118Т -1.

Далее проведено сравнение загрузки на примере молекул флуоресцентного красителя на двух МОК: МОК, полученных микрофлюидным методом, а также МОК, синтезированных сольвотермальным методом (Рисунок 3).

Продемонстрирована загрузка красителя и дистанционное высвобождение молекул красителя, при воздействии непрерывного лазерного облучения с длиной волны (532 нм). Так, продемонстрирована повышенная поверхностная дефектность МОК НКи8Т-1, полученная путем применения микрофлюидного синтеза в непрерывном потоке и проявляющаяся в увеличении адсорбции молекул красителя.

А 0.006

-£?0.004

О)

о

О 0.002 0.000

Рисунок 3 - Загрузка красителя в МОК. (А) Нелинейная аппроксимация изотермы адсорбции МОК для Су5-В8А, полученной микрофлюидным методом

синтеза (1) и сольвотермальным методом синтеза (2); (Б) Изображения клеток, инкубированных с НКи8Т-1, загруженными Су5-В8А, при конфокальной лазерной сканирующей микроскопии.

Вторая часть третьей главы посвящена синтезу перовскитов. Химически синтезированные неорганические перовскитные лазеры на основе галогенида свинца, генерирующие когерентный свет при комнатной температуре и имеющие низкий порог лазирования, стали востребованным инструментом для различных фотонных применений. Однако для изготовления высококачественных наноструктур на самой подложке важно обеспечить определенные параметры поверхности подложки и воссоздать точные условия синтеза. В разделе §3.1 рассмотрен одноэтапный метод изготовления высококачественных нанолазеров СбРЬВгз в виде суспензии, полученной путем быстрого осаждения в микрофлюидном чипе (МФЧ) [А4-А7]. Рассмотрены две геометрии

Э (1) Microfluidic synthesis ® (2) Solvothermal synthesis Langmuir fitting for (1) Langmuir fitting for (2)

f!

0.003

0.006

0.009

С (mg/ml)

0.012

микрофлюидного чипа: (а) геометрия клапана, а также (б) геометрия пористой мембраны, которые позволяют контролируемо синтезировать кристаллы перовскита (Рисунок 4).

Рисунок 4 - Дизайн геометрии МФЧ. (а) Схема конфигурации МФЧ. (Ь) Конфигурация клапана (синяя рамка). Сверху вниз: схема клапана, распределение

концентрации потока в МФЧ и вставка со схематическими изображениями полученных конструкций, (с) Конфигурация пористой мембраны (розовая рамка). Сверху вниз: изображение поверхности пористой мембраны, распределение концентрации потока в МФЧ и вставка со схематическими изображениями

полученных структур.

Изучено влияние соотношения смешиваемых в канале МФЧ прекурсоров перовскита к органическому лиганду на формирование кристаллов СбРЬВгз.

Выявлено и экспериментально показано, что примененная методика синтеза позволяет контролировать морфологию наноструктур (размер и форму) в зависимости от расхода реагентов в микроканалах (Рисунок 5).

Ю » « 90 W 90 01 01 W Р5 Tf

W ON Р5 90 Р5 90 VO 01 01 w ог vc

IPA:perovskite precursor

IPA:perovskite precursor

Рисунок 5 - Геометрические параметры частиц перовскита различной формы, полученных при различных соотношениях 1РА/прекурсор перовскита. (а)

Зависимость среднего линейного размера (ёипеаг) от соотношения 1РА к прекурсору перовскита. (Ь) Зависимость количества нитевидных наноструктур (в %) от соотношения 1РА к прекурсору перовскита.

Помимо этого, изучены структурные свойства полученных перовскитов СбРЬВгз [А5]. А также, продемонстрирована низкопороговая лазерная генерация, что подтверждает высокую кристалличность синтезированных перовскитов [А7].

Четвертая глава диссертационной работы посвящена синтезу кристаллов методом капельной микрофлюидики.

Perovskite precursors

Рисунок 6 - Синтез перовскита СбРЬВгз с помощью капельной микрофлюидики: схема используемого микрофлюидного чипа; численное моделирование

образования капель; оптическое изображение образования капель в микрофлюидном чипе; и иллюстрация влияния различных параметров (гидрофильно-липофильный баланс (НЬВ) и температура (Т °С)) на рост нанокристаллов. Масштабная шкала составляет 200 мкм.

В качестве изучаемого материала используется перовскит СбРЬВгз. Несмотря на высокий потенциал перовскитов, их стабильный, масштабируемый и воспроизводимый синтез с точным контролем размера получаемых структур (узкое распределение по размерам) является сложной задачей. В этой части диссертационной работы продемонстрирован метод синтеза полностью неорганических перовскитов из галогенида свинца с помощью капельной микрофлюидики, где сформированная капля представляет из себя герметичный микрореактор для нуклеации и последующего роста перовскитов (Рисунок 6). Для синтеза использованы две системы с различными жидкостями в качестве несущей фазы (фторированное масло и дистиллированная вода) [А8]. Продемонстрирован контролируемый синтез частиц перовскита размерами от 18 нм до 650 нм, при вариации применяемой температуры синтеза в диапазоне от 100 до 160°С с шагом

20 °С и вариации гидрофильно-липофильного баланса в диапазоне от 6 до 15 с шагом 3 (Рисунок 7).

Рисунок 7 - Структурная характеризация наночастиц перовскита СбРЬВгз. Изображения наночастиц перовскита со сканирующего электронного микроскопа (СЭМ); на вставках показаны изображения флюоресценции (ФЛ) наночастиц при

засветке УФ-лампой с длиной волны 360 нм, (а) полученных при изменении температуры синтеза (Т = 100, 120, 140, 160°С). Масштабная шкала равна 200 нм для изображений СЭМ и 2 мкм для изображений ФЛ, (б) полученных при изменении гидрофильно-липофильного баланса (ГЛБ = 6,

9, 12, 15). Масштабная шкала равна 2 мкм.

Получены оптически резонансные субмикрометровые кубоидные частицы. Измерены спектры спонтанного и стимулированного излучения в условиях окружающей среды от кубоидов СбРЬВгз микронного размера, высаженных на стеклянную подложку при засветке излучением X = 350 нм. При плотности мощности падающего излучения ниже порога генерации (Б < Бш) кубоид перовскита демонстрирует выраженный спонтанный пик с центральной длиной волны около 526 нм и шириной на полувысоте (Р\¥НМ) около 17,5 нм.

Основные результаты и выводы диссертации:

1. Показано, что синтез метал-органических каркасов типа HKUST-1 (C18H6CU3O12) в непрерывном потоке микрофлюидного чипа результирует в повышенной поверхностной дефектности МОК. Также, показано увеличение адсорбции молекул органических красителей родамина Б и цианина-5 кристаллами HKUST-1, синтезированных при использовании микрофлюидных технологий, в сравнении с адсорбцией микрокристаллов HKUST-1, полученных сольвотермальным синтезом, в 2 раза.

2. Продемонстрирован синтез перовскитов в непрерывном потоке микрофлюидного чипа, который позволяет контролировать морфологию перовскитных частиц. Впервые получены структуры перовскита в виде суспензии, что позволяет высадить их на структурированную поверхность.

3. Продемонстрирован синтез перовскитов методом капельной микрофлюидики с вариацией параметров синтеза (температуры и ГЛБ). Впервые продемонстрирована вариация размера структур перовскита CsPbBr3 при изменении гидрофильно-липофильного баланса поверхностно-активных веществ при синтезе методом капельной микрофлюидики.

Основные публикации по теме диссертации:

[А1] Koryakina, I., Kuznetsova, D. S., Zuev, D. A., Milichko, V. A., Timin, A. S., & Zyuzin, M. V. Optically responsive delivery platforms: From the design considerations to biomedical applications //Nanophotonics. - 2020. - Т. 9. - №. 1. - С. 39-74.

[А2] Koryakina I., Afonicheva P., Arabuli K., Timin A., Evstrapov A. & Zyuzin M. V. Microfluidic synthesis of optically responsive materials for nano-and biophotonics //Advances in Colloid and Interface Science. - 2021. - Т. 298. - С. 102548.

[А3] Koryakina I., Bachinin S., Gerasimova E., Timofeeva M., Shipilovskikh S., Bukatin A., Timin A., Milichko V. & Zyuzin M.V. Microfluidic synthesis of metal-

organic framework crystals with surface defects for enhanced molecular loading //Chemical Engineering Journal. - 2023. - T. 452. - C. 139450. [A4] Koryakina I. G., Naumochkin M., Markina D. I., Khubezhov S. A., Pushkarev A. P., Evstrapov A. A. & Zyuzin M. V. Single-Step Microfluidic Synthesis of Halide Perovskite Nanolasers in Suspension //Chemistry of Materials. - 2021. - T. 33. - №2. 8. - C. 2777-2784.

[A5] Koryakina, I., Pushkarev, A., Makarov, S. V., Zyuzin, M. V., & Evstrapov, A. A. Synthesis of perovskite nanoparticles in microfluidic chips //Journal of Physics: Conference Series. - IOP Publishing, 2020. - T. 1461. - №. 1. - C. 012071. [A6] Koryakina, I., Markina, D., Evstrapov, A., Pushkarev, A., Makarov, S., & Zyuzin, M. V. Microfluidics-based synthesis of lead cesium bromide perovskite microcrystals //AIP Conference Proceedings. - AIP Publishing, 2020. - T. 2300. -№. 1.

[A7] Koryakina I., Bikmetova S., Arabuli K., Evstrapov A., Pushkarev A., Makarov S. & Zyuzin M. V. Continuous-Flow Synthesis of Perovskite Particles for Optical Application //Journal of Physics: Conference Series. - IOP Publishing, 2021. - T. 2015. - №. 1. - C. 012072. [A8] I. Koryakina, S. Bikmetova, D. Khmelevskaya, D. Markina, A. Kuleshova, L. Logunov, A.Timin, A. Pushkarev, S. Makarov, Mikhail V. Zyuzin. Droplet Microfluidic Synthesis of Halide Perovskites Affords Upconversion Lasing in Mie-Resonant Cuboids //ACS Applied Nano Materials. - 2023. - T. 6. - №№. 6. - C. 43704378.

26

Synopsis

General thesis summary

Relevance of the chosen topic

Optically sensitive materials are a class of materials that exhibit unique properties under the influence of radiation. Depending on their nature, such materials have different optical functionality. A remarkable representative of the class of optically sensitive materials are metal-organic frameworks (MOFs). MOFs are organic-inorganic hybrid crystalline porous materials that have a large internal surface area, homogeneous pore structure, adjustable porosity and chemical functionality. These parameters allow us to successfully vary the structure of the MOF and its functionality, which, in turn, enables development of selectively absorbing materials (e.g., selective absorption of gas molecules [109; 127], dye molecules [99] or drug delivery [81; 88; 134]). This feature can be improved by optical sensitivity of MOFs, for example, initialization of cargo release when MOF is exposed to the optical radiation of a certain wavelength. The most common methods of MOF synthesis are chemical methods - solvothermal and microwave.

Another optically sensitive materials with a broad application in photonics and optoelectronics are halide perovskites CsPbX3 (where X = Cl, Br and I). Nanostructured perovskite materials are of interest due to such properties as tunable band gap, light emission in the entire visible spectrum, narrow luminescence line width, as well as multiphoton absorption. In addition, lead halide perovskites of various shapes (spheres, cubes, wires, disks, rings, etc.) can serve as resonant cavities with an active optical medium for the development of microscale photonic devices [71]. Conventional approaches to the synthesis of CsPbX3 (X = Cl, Br and I) include laser lithography [90], chemical vapor deposition in an inert atmosphere [31], as well as hot-injection, which is time-consuming, sensitive to environmental conditions and requires expensive equipment.

The abovementioned methods of synthesis of both perovskites and MOFs operate with macroparameters, which affects the quality of the synthesizing materials. It should

be emphasized that the physicochemical properties of materials (phase, size, morphology and composition) are mainly determined by the procedure of their synthesis, which usually includes several stages when using conventional synthesis methods. These methods are not always reproducible due to insufficient control of heat and mass transfer.

Список литературы диссертационного исследования кандидат наук Корякина Ирина Георгиевна, 2023 год

ro >■

'in

c

(U

. 1 1 I #27093 Cubic

JL J A J » 11 1 * A A

1 j 1 1 J » A 500K A

i 1 1 . 1 A 450K A

1 J 1 À 1 A 425K ^ A

i 1 I, A A À A I 410K

i ' 1 A A 400K A

1 1 A #109295 tetragonal Jl À . .. . i

< 1 .J I A A à 370K

X - 1 J A A I 360K l A

I i A K 350K A

i i . » . . J I k A I 300K A.

I #97851 orthorhombic

A j —'—i—■—i— .....i -'-1---r 1 A —'—i— A 1 -i—î—i— -1- ->-1-'—

u lû

U

fO

c o

ÛO ro

<u

Î

u

1Q

E o

e

O

10 15 20 25

30 35 40 45 2Theta(deg)

50 55 60

Figure 4 - Powder X-ray diffraction patterns showing the phase transition of CsPbBr3 perovskite depending on the temperature [68].

1.1.2. Conventional synthesis methods of perovskites

Nowadays there are a plethora of various synthesis methods that enables the efficient outcome [33; 57].

In order to grow high-quality perovskite single crystals. There were attempts to fabricate perovskite nano- and microstructures using the chemical vapor deposition method (CVD) [11-14]. Other variations of the CVD method can be hybrid chemical vapor deposition (H-CVD), vapor assisted solution processing (VASP), in situ tubular chemical vapour deposition (ITCVD), low pressure CVD (LP-CVD). The main idea of these methods is the same and consist of growing the perovskite structure by deposing

the sublimed organo-halide material onto the substrate with pre-deposited metal halide precursor layer under stable isobaric-isothermal conditions (Figure 5).

Figure 5 - Scheme of CVD method of perovskite synthesis [118].

The temperature of the deposition as well as the pressure can be defined, which enables high control over synthesis conditions. But it should be taken into account that this method requires costly equipment.

It is found that the formation of the perovskite occurs due either mass transport or diffusion-limited reactions, depending on the temperature of the organo-halide component sublimation zone [118].

Another method for the perovskite synthesis is the hot-injection method. The formation of perovskite is initiated when one of the precursors (Cs-oleate) is injected into a pre-heated to 120oC solution of dissolved halide (PbBr2) (Figure 6) [72; 85]. Kinetics control of perovskite formation is achieved by varying the concentration of the ligands (oleic acid and oleylamine) and the ratio of the precursors [56]. The formation and growth mechanism of perovskite in this case can be described with LaMer model [100].

One more widely employed method for the perovskite synthesis is the ligand-assited represipitation (LARP). This technique combines the mixture of the good (e.g. AA-dimethylformamide (DMF), tetrahydrofuran (THF), and dimethyl sulfoxide (DMSO)) and poor solvent (e.g. toluene or hexane) to provide self-quenching of the perovskite growth reaction [69].

1.1.3. Metal-organic frameworks

Metal-organic frameworks (MOFs) are crystalline hybrid materials that consist of metal ions (or metal clusters) nods connected via organic linkers (Figure 7).

Figure 7 - Scheme of MOFs components that form porous, crystalline structure [30].

Variation of components allows to diversify the structure and its physical and chemical properties. The intense research of the past decades led to more than 100 000 MOFs being reported [37; 103].

Figure 8 - The 3D structures of representative MOFs [103].

Beside structural tunability (Figure 8), MOFs exhibit high surface area, chirality, luminescence, high crystallinity, and porosity [59]. Well-defined pores form regular crystalline lattices with interesting properties.

The most prominent applications of MOFs are gas separation and storage, data storage, chemical sensing, cargo delivery, catalysis, etc. [97].

1.1.4. Conventional synthesis methods of MOFs (tunability of intrinsic properties)

The final properties of the MOFs are influenced not only by the initially selected components, but also by the synthesis condition. For example, different morphology of MOFs' particles can influence the adsorption and separation of guest molecules and gases, and catalytic properties. Small crystals may also be employed for the formation of membranes using a seeded growth procedure.

Nanocrystalline, nontoxic MOFs with high loading capacities are also envisioned for biomedical application. For these objectives, sonochemical and microwave-assisted methods have been applied. On the other hand, larger single crystals (in the 100 ^m range) are necessary to routinely determine the structure and to establish structure property relationships.

Examples of most commonly applied synthesis methods of MOFs are solvothermal, microwave-assisted, ultrasonication, electrochemical, mechanochemical, ionothermal, and dry gel conversion synthesis methods [89].

The principle of solvothermal synthesis is in mixing MOFs precursors under constant temperature in the organic solution [59]. The definition of the solvothermal method is given by Rabenau [34] and states as follows "reactions taking place in closed vessels under autogenous pressure above the boiling point of the solvent". Solvothermal synthesis is usually based on electric heating in vials, tubes, or autoclave, and requires high temperatures in the range of 100-180oC during a prolonged period of time, e.g., 1296 h.

In case of microwave-assisted synthesis the interaction of electromagnetic waves with mobile electric charges occurs. These can be polar solvent molecules/ions in a solution or electrons/ions in a solid. In the solid, an electric current is formed and heating is due to electric resistance of the solid. In solution, polar molecules try to align themselves in an electromagnetic field and in an oscillating field so that the molecules change their orientations permanently. Thus, applying the appropriate frequency, collision between the molecules will take place, which leads to an increase in kinetic energy, i.e., temperature, of the system. Due to the direct interaction of the radiation with the solution/reactants, MW-assisted heating presents a very energy efficient method of heating. Thus, high heating rates and homogeneous heating throughout the sample is possible. Attention must be paid to the choice of appropriate solvents and selective energy input, since starting materials may strongly interact with the MW radiation.

Depending on the synthesis conditions, the duration of the reaction can vary from 10 to 60 min. In the ultrasound synthesis approach, the acceleration of the reaction time (around 30 min) is achieved by extreme local heating up to 5000 K [130]. MOFs synthesis techniques are comprehensively described in the works [34].

All above mentioned synthesis methods possess long synthesis time. To reduce the synthesis time of optically responsive NPs microfluidics can be applied. Microfluidics provides an outstanding control over parameters of nanomaterials synthesis, resulting in NPs with predictable sizes, shapes, and particle size distributions. Microfluidic technologies enable reproducibility of the synthesis route and scaling up the resulting product. Due to the excellent heat and mass transfer in microchannels of a microfluidic chip (MFC), synthesis time can be reduced [99].

1.2. Formation mechanisms of nanoparticles

To understand the formation process of nanomaterials, we should consider several models (e.g., LaMer model, Ostwald Ripening and Digestive Ripening, Finke-Watzky mechanism) developed for controlling the morphology and size of NPs during their

synthesis [83; 131; 136]. These models describe the NPs' formation mechanisms using the conventional wet chemistry methods.

LaMer model is considered as the most widespread theory explaining the process of particle formation in the solution. It usually includes four steps, such as (i) supersaturated solution formation, (ii) nucleation, (iii) growth and (iv) agglomeration (Figure 9). At the first step, solute concentration increases up to the maximum (the so-called critical solute concentration). Next, the nucleation process occurs, in which the components of the reaction undergo a "burst-nucleation", and after that, the concentration of free components in the solution is significantly reduced. The low concentration of components complicates the nucleation process, therefore, the particle growth begins as the components diffuse through the solution [83; 136]. Since the growth of the particles can lead to the particle aggregation, it should be carefully controlled to form a stable particle sol. The sol-gel technique is a good example demonstrating the process of particle formation, nucleation and growth [136]. According to the LaMer method, it is vital to separate the nucleation and growth processes in order to form monodisperse particles.

Figure 9 - Formation mechanisms of NPs. (A) The LaMer model: a schematic diagram of particle formation with the corresponding SEM images of Au mesoflowers

synthesized at 5 s, 1 min, 3.5 min, and 10 min, respectively [66]. (B) Ostwald ripening and digestive ripening with the corresponding SEM images of the seed growth of Au

NPs [7].

The Ostwald Ripening mechanism of particle growth is based on the change in the solubility of particles depending on their size [20]. Due to the high solubility and high surface energy of small particles in solution, they can be redissolved, and then, further particle growth is achieved [136].

The Digestive Ripening is a process that is opposite to the Ostwald Ripening. The Digestive Ripening method of nanomaterials' formation includes three steps: (i) re-dispersing large particles into small particles by the addition of ligands; (ii) extracting the products from the synthetic mixture; (iii) formation of monodisperse NPs by the heating process.

Figure 10 - Formation mechanisms of NPs. (A) Finke-Watzky mechanism of NPs transformation with the corresponding TEM images of Au NPs (KBr was used as additive). Scale bar is 250 nm [45]. (B) Coalescence attachment mechanism of NPs'

transformation with the corresponding TEM images of two attached Au NPs during the

fusion process [5].

The Finke-Watzky mechanism of NPs formation includes two main processes: (i) continuous nucleation, and (ii) autocatalytic growth. This mechanistic model is based on the concept of continuous nucleation followed by the autocatalytic surface growth (Figure 10) [22]. The nucleation process is slow, while the autocatalytic growth of particles occurs rapidly.

Coalescence and Orientated Attachments are similar processes of NPs growth. Once the NPs are attached to each other as building blocks, they agglomerate into a single NP. Here, the reduction of the surface energy of NPs is the driving force, since the total surface area of the united particle is lower than the sum of areas of each individual particle before their coalescence [76]. The Orientated Attachments usually occur in two pathways: (i) NPs can be united by three-dimensional (3D) rotation, which decreases interparticle misalignment, defeating the dislocation of particle-particle barriers; (ii) atom-by-atom reorientation via grain boundary migration.

Therefore, microfluidic synthesis can be used to control each step or mechanism in various models of NPs formation. Indeed, this approach was already applied for the large-scale continuous production of hybrid NPs with definite size, shape, structure and properties [131].

1.3. Overview of Microfluidics

Microfluidics operates with micro- and nano-volume liquids, which makes it possible to increase the accuracy of the reaction by varying parameters such as the ratio of reagents and reaction time. The high rate of heat and mass transfer in the channels of the microfluidic chip significantly accelerates the time of the synthesis of materials.

In microfluidics laminar and turbulent flow depending on the nature of the fluid movement are distinguished. Laminar flow refers to the liquid moving in layers without any mixing, with interaction between different liquids occurring only at their interface

through diffusion. Turbulent flow, on the other hand, is characterized by the formation of vortices and chaotic changes in the velocity and pressure of the fluid flow. A dimensionless characteristic number, the Reynolds number Re, is used to distinguish these two flows:

Re = (1)

%

where v is the velocity field, p is density, L is a characteristic size of the channel, which can be defined as L = 4A/P , where A is the cross-sectional area of the channel, P is the perimeter, p is the viscosity.

At low Re the flow is laminar (in microfluidics Re is usually less than 1), whilst at high Re it is considered turbulent. Due to the small scale (L parameter is typically in the range of mm or ^m), Re in microfluidics is usually less than 1. This results in laminar flow in microfluidic channels. Laminar flow is simpler; therefore, it is easier to control the flow, and predict the behavior of fluids. However, for some applications, turbulent flow is required. For example, for material synthesis, controlled mixing of two or more reactants is necessary. The mixing type describe the interaction between the reagents, which affects crystal growth and ultimately determines the properties of the resulting materials.

Figure 11 - Laminar flow scheme. (A) Velocity profile. (B) Reagents mixing

Further, advantages and disadvantages of each flow type for the synthesis of materials are considered. Laminar flow creates continuous flow (Figure 11) that is characterized as a flow with no breakings of the continuity or uninterrupted flow. In the case when mixing of two or more different liquids is considered the mixing occurs only when reagents mutually diffuse on the interphases [67]. This enables to predict the outcome with higher precision, and enables better control over synthesis condition. It is also should be stressed that continuous flow defines a pronounced gradient of diffusion which determines the wide dispersion of the reaction outcome (e.g., product sizes, product shape).

On the contrary, turbulent flow creates twirls and chaotic mixing (Figure 12); therefore, the homogenous mixing is reached faster. However, in microfluidics, this type of flow requires internal or external stimuli. Such stimuli can be classified as passive type of mixers and active type of mixers.

Passive mixing uses various modifications of the MFC channel's geometry to enhance mixing without applying external energy (except pressure used to drive the fluid flows). Geometry of an MFC can involve spiral or zigzag channels, embedded barriers and others, which can also be combined with each other. The main advantage of the passive mixing approach is the absence of additional complex parts integrated in an MFC

for mixing. In the work [50] the enhancement of mixing in the droplet microfluidics is achieved by changing the angle of the continuous flow supply. Therefore, droplet microfluidics can be considered as a passive method of mixing enhancement in the microchannels. However, sometimes the application of a sophisticated topology is not enough, therefore, to ensure an effective mixing of fluids with high viscosity, the introduction of active components is relevant (Figure 13A).

Figure 13 - Active and passive mixing approaches. (A) Passive mixing in the MFC channel. Mass fraction contour of the "T" channels with (i) 0, (ii) 1, (iii) 5, (iv) 10-period mixing unit at a flow rate of 20 ^L/s through the outlet channel. The arrows indicate the flow direction [102]. (B) Active thermal mixer. Schematic view of the MFC channel with heaters placed inside [36]. (C) Active acoustic mixer. Scheme of the flow distribution, and behaviour of fluorescein particles in the presence of acoustic waves [117]. (D) Active pressure-driven mixer. Schematic view of the actuators' location in

the MFC chamber [42].

Active mixing involves external stimuli [123] such as pressure, acoustic field, electric field, magnetic field, electrokinetics, thermal energy, and others (Figure 13B,C,D) [55; 98], which induce the mutual diffusion of flows. Pressure-driven mixing that can be realized by introduction of micropumps. The mixing is based on peristaltic movements enabling active mixing of fluids. Acoustically induced microflows are also widely used for mixing of different types of fluids in microfluidic devices. Acoustic-driven micromixers employ bulk acoustic waves, surface acoustic waves and acoustically vibrating microstructures to transfer acoustic energy into liquids and improve their mixing. Air bubbles can also be used as an active element of an acoustic driven mixer. For this, the air bubbles are introduced into a channel, and acoustic waves are applied, affecting the bubbles in the fluid. Thus, the bubbles are displaced, causing radial and translational oscillations, which lead to the flow pulsation. Therefore, the bubbles become additional acoustic transmitters and amplify the effect of acoustic waves. The acoustically-excited bubbles can be applied for the development of microvortices and reduction of mixing time to a few seconds. Thermal energy also can be used to increase the mixing efficiency by enhancing the diffusion of two flows with thermal bubbles or thermal effects. Another kind of active mixer is based on magnetic field-driven mixing. Such a mixing can be realized by introducing magnetic NPs or moving structures (a microstirrer or a rotor) into the MFC's channel and inducing their motion in the fluid by a permanent magnet, an electromagnet or integrated electrodes. This in turn leads to the formation of turbulent flows and the diffusion of liquids.

Despite the obvious benefits (rapid and homogenous mixing) of active mixers, they possess some disadvantages: the whole microfluidic system becomes complex with addition of extra parts, which complicates the fabrication process. Beside from that, it should be noted that active mixer affects the kinetics of a reaction as well.

Successful implementation of the turbulent flow advantages of fast mixing for the material synthesis is realized in the concept of droplet microfluidics. Droplet microfluidics study controllable emulsification, or droplet formation, with homogeneous

size and high-frequency of droplet formation (up to 10 000 droplets per unit of time) [16; 74].

In microfluidic devices, droplets or segments can be produced using passive or active approaches. Passive droplets/segments formation methods are defined by the geometry of an MFC based on different physical mechanisms acting when the flow of the dispersed phase breaks up into drops in: (i) coaxial flows (co-flowing stream), (ii) elongation of deformed flows in devices with flow focusing (flow focusing), (iii) intersecting flows in a T-shaped device (T-injector) [64] (Figure 14).

Figure 14 - Types of droplets/segments formation. (A) A fundamental difference between the drops and segments in an MFC channel (shown in T-injector geometry).

(B) Passive droplets/segments formation geometries: (i) co-flowing stream, (ii) flow

focusing, (iii) T-injector [64].

In the coaxial-flow (co-flow) geometry, generation of drops occurs when a capillary with dispersed phase is placed inside a channel with continuous phase [64; 114]. In the case of a flow-focusing MFC geometry, drops or segments are produced by the hydrodynamic focusing (shearing) of the dispersed phase by the continuous phase. T-injector consists of two channels: an inlet channel containing a dispersed phase is perpendicular to the main channel containing a continuous phase. Drops/segments in the T-injector MFC configuration are formed after the dispersed phase enters the main channel, and the continuous phase takes it downstream. In the cases of coaxial and flow-focusing droplet generation, it is crucial to make sure that the dispersed phase does not wet the walls (see §1.3.2.3). Surface wetting affects the interfacial tension, which is crucial for the droplet formation.

For each MFC geometry, generation of droplets/segments can occur in three regimes: dripping, jetting, and stable. In the case of the dripping regime, drops detach from the dispersed phase in the capillary and are carried downstream by the continuous outer flow. In the jetting regime, dispersed phase forms a jet (thread) and separates to drops when destabilization occurs. Stable co-flow regime does not imply drop formation, since the system is stable.

At the interface of liquid media, the following forces affect the formation of droplets/segments: the viscous force of shear stress, the force of surface tension, and the resistance force to the continuous-phase flow (compression force). The formation of the drop/segment can be regulated by the ratio of the flow rates between two immiscible liquids.

Reagent distribution

a.u

Figure 15 - Numerical simulation of the droplet formation process and the concentration distribution during droplet formation in the asymmetric (A) and symmetric (B) droplet generators. The scale bar is 30 ^m [50].

Each drop represents an isolated microreactor and improves mixing of reagents to a more uniform profile. Small droplets enhance mixing due to their size and the presence of vortex flows within them (Figure 15). Additionally, droplets can be subjected to specific physical effects, such as temperature.

1.3.1. Theory of microfluidics

Microfluidics is a part of hydrodynamics, therefore most of the physical laws of hydrodynamics are applicable in microfluidics. Although, due to the smaller scale considered in case of microfluidics some of macro scale parameters can be neglected (e.g.

gravity force), and new micro scale forces are taken into consideration. Such forces, that prevail at micro scale are the physical characteristics of a liquid, such as its density and viscosity. Those parameters affect flow behavior. The density of a liquid depends on the mass of an individual molecule and the number of the molecules that occupy a unit of volume, which determines its inertia or resistance to an accelerating force [133].

The ability of a fluid to deform under a shear force is a defining characteristic of liquids and gases. The behavior and interactions of these different states of matter are determined by the intermolecular forces between their constituent molecules and can be described by Navier-Stokes equations.

1.3.1.1. Navier-Stokes equation

The fluid behavior can be modeled with the Navier-Stokes equations, that describes the fluids motion based on the internal and external forces. In the case of considering the fluid flow at the micro and nano scale the Navier-Stokes is written as given further: p(d&v + (v • V)v) = -Vp + ^V2v + pg + PelE (2)

where v is the velocity field, V is the nabla operator, p is the pressure, n is the viscosity, pg is the gravitational force (in terms of the density p and the acceleration of gravity g), and pelE is the electrical force (in terms of the charge density pel of the fluid and the external electric field E).

Due to the scale, laminar flow prevails in the case of microfluidics. The Navier-Stokes equation with corrections for each case is used to describe the flow physics.

Together with the equation (2), the continuity equation is solved:

+ V • (pu) = 0 (3)

The conservation of mass is represented by the continuity equation, whereas the conservation of momentum is represented by the Navier-Stokes equations.

1.3.1.2. Dimensionless characteristic numbers

The flow in the microchannels can be characterized by dimensionless numbers [100]. One of the most commonly used dimensionless numbers is the Reynolds number (Re), which represents the ratio between the inertial and viscous forces.

Re = (1)

%

where v is the velocity field, p is density, L is a characteristic size of the channel, which can be defined as L = 4A/P , where A is the cross-sectional area of the channel, P is the perimeter, p is the viscosity.

Small values of Re indicate laminar flows, in which viscous forces dominate and determine the smooth movement of a liquid. It is assumed that in the microfluidic systems, Re < (1^10). At high Re, the inertial forces significantly affect the flow and, therefore, turbulence occurs.

Another dimensionless number is the Capillary number (Ca), which estimates the relative strength of the shear stress compared to the surface tension.

Ca = %+ (4)

where y is the interfacial tension.

In case of droplet microfluidics, the lower the surface tension is, the easier it is to generate fine droplets. At relatively high values of Ca, viscous forces dominate, which leads to the deformation of the droplets by the flow and the appearance of droplets with asymmetric shapes when the flow rate remains without changes [129].

1.3.1.3. Diffusion

In microfluidic system with Re<<1 the mixing process is defined by intermolecular diffusion that happens via Brownian motion. Therefore, diffusion (D) can be defined as the irregular motion of microscopic particles suspended in a fluid due to collisions with the surrounding fluid molecules. D is directly proportional to the temperature of the fluid (T) and inversely proportional to the viscosity of the fluid (r¡) and the diameter of the particle (a) [18; 82]:

D = — (5)

6 nr¡a

where k is Boltzmann's constant.

One of the main requirements for the diffusion is the concentrations gradient. Mixing in this case can be estimated with diffusive flux (6):

J = -D (6)

where J is the diffusive flux (mol/(m2s)), D is the diffusion coefficient (m2/s), and C is the concentration of drug in the reservoir (mol/m3).

To enhance mixing in microfluidics various micromixers can be incorporated within microfluidic systems to increase mixing efficiency. Generally, they are divided to active and passive mixers, as described in §1.3.

1.3.1.4. Hydrophilic-lipophilic balance

In droplet microfluidics it is essential to ensure the stability of the formed droplets. To enable stability of droplets, surfactants (abbreviation for surface-active agents) can be applied [8; 53].

Surfactants are amphiphilic molecules that reduce the interfacial tension between the reagents of continuous and dispersed phases. The choice of an appropriate surfactant is crucial. To facilitate the process a hydrophilic-lipophilic balance (HLB) concept was first introduced by Griffin in 1949 and further elaborated on in 1954. Griffin suggested that the degree of hydrophilicity or lipophilicity of a surfactant, known as its hydrophilic-lipophilic balance, is determined by analyzing the molecular weight percentages of its hydrophilic and lipophilic components. These methods were initially designed for nonionic surfactants and rely on the sizes of the hydrophobic and hydrophilic parts of the surfactant molecules. HLB numbers can predict the surfactant's ability to stabilize oil-in-water (o/w) or water-in-oil (w/o) emulsions. If the HLB value is between 4 and 6, it generally stabilizes w/o emulsions, whereas values between 8 and 18 stabilize o/w emulsions[41; 112]. The literature contains tabulated HLB numbers for many surfactants, and the HLB numbers of surfactant mixtures can be calculated from their respective weight proportions using equations.

The percentage ratio between surfactants was calculated by the equation:

%(A) =

%(B) = 100- %(A)

(8)

where %(A) is the percentage of surfactant A, %(B) is the percentage of surfactant B, X is the required HLB value.

Further the mass of the surfactant was calculated by the equation:

where M is the mass of the surfactant mixture.

1.3.2. Materials of microfluidics

The advancement of microfluidics-based technologies in various scientific fields heavily relies on the combined progress of materials and microfluidic platforms. The utilization of new materials has broadened the scope of microfluidics applications, while microfluidic systems have proven to be reliable and adaptable platforms for producing materials with precise physicochemical characteristics.

A variety of materials, including silicon, quartz/fused silica, glass, ceramics, polymers, and metals, have been utilized in constructing microfluidic devices with diverse functionalities [91].

Materials not only provide mechanical support and define the structures of embedded microfluidic networks, but also have a crucial role in determining the specific properties, overall performance, and range of applicability of microdevices [137]. The material can act as a medium for heat transfer, electrical conduction, transmission of light, or transduction of mechanical force, while the surfaces forming channel walls interact with the fluid medium and influence its constitution and physical behavior. These material functions stem from their intrinsic physical and chemical properties, which can be controlled through various techniques. Table 1 outlines common materials used for

(9)

(10)

m(B) = M - m(A)

1.3.2.1. Overview of materials

microfluidic device fabrication and their relevant properties impacting device performance.

Table 1. Materials properties [91].

Material property Si (single-crystal) Glass SiO2 PDMS PMMA Polycarbonate

Coefficient of thermal expansion (10°C-1) 2.6 0.55 0.55 310 55 70.2

Thermal conductivity at 300K (Wcm-1K-1) 1.57 0.011 0.014 0.0018 0.002 0.002

>70% optical transmittance (nm) >700 >350 >350 400700 400700 400-700

Maximum processing temperature (°C) 1315 550600 1700 ~150 ~100 ~100

Bulk resistivity (^Q) 2.3*1011 >10 >10 >10 >10 >10

Dielectric strength (10 Vcm-1) 3 5-10 2-3 2.1 0.170.19 0.39

Water contact angle, advancing (degrees) 110 20-35 ~30 ~110 60-75 78

When considering the material of the future MFC several material properties usually are taken into account, depending on the use conditions.

Glass microfluidic chips are excellent for the synthesis of materials. The glass is transparent in the UV and visible range, making it ideal for optical detection and imaging. Additionally, its chemical and thermal resistance make glass suitable for use in chemical-reaction devices.

1.3.2.2. Methods of MFC fabrication

Typically, microfluidic devices are built by creating open microchannels or features in a flat substrate through dry or wet etching, mechanical machining, molding, or casting. This substrate is then bonded to another flat plate with predrilled access holes, resulting in an enclosed microchannel network with desired dimensions and configuration. These materials serve as the building blocks or substrates for various types of microfluidic devices.

It is worth mentioning that in micro-dimensional spaces, the liquid interacts with the surface of the MFC channels, and therefore, with the MFC material. Therefore, the properties of the MFC material have a significant impact on the flow movement and processes that occur when the fluid interacts with the surfaces of microchannels. For instance, roughness of an MFC channel can affect the liquid heating up more than a smooth one, since a large surface area is applied for heat transfer. Naturally, the smaller the space for fluid movement, the more significant the influence of the surface properties on the fluid.

To mitigate MFC surface interaction with the liquid in its channels, the surface is coated with a material.

When considered droplet formation (emulsification) in the channels of an MFC it is important to mention the wettability of the MFC walls. For reproducible emulsification of highly monodisperse drops it is crucial that the dispersed phase is not wetting the MFC walls. For this purpose, the MFC can be treated with additional liquids that have affinity to the continuous phase.

1.3.2.3. Wettability of walls

gas

contact single

liquid

solid

Figure 16 - Contact angle for wettability characterization

The wetting behavior of an MFC channel's wall is characterized by the contact angel (Figure 16). The contact angle of a solid surface is determined by the interfacial energies between the solid and liquid. When a liquid droplet is placed on a solid surface, tensions arise from the three interfaces: liquid-gas, solid-gas, and solid-liquid. Young's theory states that equilibrium is achieved when the horizontal components of the interfacial energies are equal.

Yig cos6=Ysg — Ysi (11)

where ylg, ysg, and ysl are interfacial tensions of liquid-gas, solid-gas, and solidliquid respectively. When contact angle is equal 0o the surface is perfectly wetted, in case of contact angle is between 0o and 90o the surface is highly wetted, whereas contact angle more then 90o characterizes nonwetting surface.

1.4. Microfluidics for materials synthesis

Microfluidic approach affords NPs synthesis with predictable sizes, shapes, and particle size distributions. Additionally, microfluidic technologies enable reproducibility of the synthesis route and scaling up the resulting product. Due to the excellent heat and mass transfer in microchannels of an MFC, synthesis time can be reduced [104; 121]. Apart from synthesis, MFCs can be endowed with some additional functionalities, such as in situ detection of the obtained NPs. Therefore, it is possible to combine all the synthetic procedures (incl. synthesis, purification, detection) in a single microfluidic platform for the automatization of the NPs synthesis.

In addition, when using a microfluidic chip, in situ detection of synthesized structures becomes possible, which enables to evaluate and adjust the synthesis conditions in situ. Due to these characteristics, the microfluidic synthesis method makes possible the synthesis of optically responsive materials with high level of crystallinity and monodispersity.

1.4.1. Perovskite synthesis via microfluidics

Nowadays, perovskite nano- and microparticles are of significant interest due to their outstanding optical and electrical properties enabling their application as lasers, solar cells, photodetectors, and light-emitting diodes (LEDs) [49; 51; 71; 108]. Nonetheless, cavities of perovskites of various shapes can serve as an active optical medium of microscale photonic devices due to their outstanding lasing properties [32; 39; 47; 107]. Conventional synthesis approaches are chemical vapor deposition and hot-injection methods. The synthesis duration can vary from 80 to 120 min.

Zhang et al. reported on a rapid synthesis of CsPbBr3 nanowires (NWs) in a continuous-flow microfluidic device [115]. Perovskite precursors were mixed in a controlled manner, and the time of a reaction was controlled by the flow rate. The obtained perovskite morphologies and their optical performances depended on the flow rates and temperature of a reaction. The authors demonstrate that the developed microfluidic approach promoted the alignment of elongated CsPbBr3 NWs, while conventional batch synthesis resulted in irregular nanorods under the same conditions. In a recent work, a novel microfluidic synthesis approach was proposed for the synthesis of high-quality CsPbBr3 NWs [32]. This method enabled transferring the obtained nanolasers onto any arbitrary surface, which is important for the fabrication of photonic devices with high throughput and excellent control over geometrical parameters. The developed microfluidic chip consisted of two channels (for perovskite precursor and isopropyl alcohol) with a valve or a porous membrane in-between. By adjusting the velocity rate in channels, CsPbBr3 particles with different geometries (cubes, NWs, rods, and others) were obtained. The obtained perovskite NWs supported lasing at room temperature in the visible range. Perovskite nanocrystals (NCs) can be also synthesized on a microfluidic platform. Lin et al. developed a microfluidic device that realized coaxial flow in a stable co-flow regime for the synthesis of all-inorganic perovskite nanocrystals [58]. The physical mixing of perovskite precursors and start of chemical reaction was initiated in the inner flow, whereas the outer fluid carried the growing NCs to the outlet. The whole growing procedure was monitored with an optical fiber absorption

spectrometer and an optical fiber fluorescence spectrometer. Microfluidically obtained perovskite NCs were further applied for the fabrication of green and white LEDs. This group applied the same technology for the templated synthesis of perovskite NCs in order to improve their stability against temperature, light and water [126]. For this, microfluidically synthesized co-flowing perovskite NCs mixed with a hydrophobic photocurable precursor were divided into droplets by the continuous phase (sodium dodecyl sulfate). Afterwards, the obtained droplets were irradiated with UV light with their following solidification. The obtained green-emitting microspheres with encapsulated perovskite NCs showed stability in water up to 150 h.

It is worth noting that not only fluids can be applied as a carrier phase to realize segments in a microfluidic system, but also gases. Abdel-Latif et al. applied an inert gas (argon) for the formation of segments during the synthesis of lead halide NCs. Authors systematically studied the influence of ligand composition and halide salt source on the composition and optical performances of the obtained NCs using in situ optical spectroscopy [65].

Despite quite successful using of the developed conventional methods for obtaining perovskites, microfluidics synthesis approach makes the process much faster (several seconds/milliseconds) and reduces the synthesis temperature.

Droplet synthesis allows one to obtain homogeneous perovskites crystals of similar morphology (e.g., NWs) compared with the conventional one due to smaller concentration gradients in one reaction chamber (one drop), where it is possible to create the same conditions for synthesis in each drop. Furthermore, synthesis of some perovskite materials requires the usage of harmful reagents, but microfluidics can automate the process of particles fabrication and allows conducting synthesis in sealed chips without air contact making the process environmentally friendly. However, some widely used materials for MFC chips (e.g., PDMS) can be not tolerable to a number of chemicals (isopropyl alcohol, dimethyl sulfoxide), which results in its swelling and destroying the chip geometry.

1.4.2. MOF synthesis using microfluidics

Metal-organic frameworks (MOFs) are a new class of crystalline porous materials that has gained a lot of interest in the last few years. The unique structural property of MOFs is their extremely high porosity and, thus, increased internal surface area that can be utilized in such applications as data storage [119; 125], gas adsorption [109; 127], drug delivery [81; 88; 134], sensing [88] and others [97; 110].0ptical properties of MOFs include optical absorption [106], PL, structural changes, and photocatalyst [26]. For instance, recently, optical switching of stimuli-responsive and photochromic-free HKUST-1 was demonstrated [110]. The main mechanism of this optical switching was explained by the dehydration-reversible structural shrinking of the MOF cage. Exposure to the NIR light led to a blue shift in the MOFs' absorption band spectrum.

To achieve fast and monodisperse synthesis of MOFs with controllable porosity of the resulting structures, the microfluidic synthesis approach can also be applied. Faustini et al. [23] demonstrated synthesis of a number of MOFs such as HKUST-1, MOF-5, IRMOF-3, and UiO-66, and beside that, Co3BTC2@Ni3BTC2, MOF-5@diCH3-MOF-5 and Fe3O4@ZIF-8 core-shell MOF crystals, which were synthesized using droplet microfluidics with T-injector design. For HKUST-1, MOF-5, and UiO-66 synthesis, the precursor solutions were prepared by mixing metal salts and organic ligands into a homogeneous solution with solvents (N,N-dimethylformamide (DMF), ethanol (EtOH) and deionized water (H2O)). To obtain drops, silicone oil was used as a continuous phase, while the whole system was heated up to 90°C for 1-12 min. By varying the reaction time, MOFs particles with different sizes were obtained. Segmented flow microfluidic synthesis approach was applied for the fabrication of MIL-88B (Fe-MIL-88B-NH2, Fe-MIL-88B, and Fe-MIL-88BBr) in the work [2]. In this case, the authors used a system of connected capillaries consisting of two intersections (Figure 9A). The segments were formed at the 2nd intersection, while at the 1 st the metal and ligand solutions were mixed. Cui et al. [101] synthesized monodisperse porous single-crystalline MOFs microcubes applying the droplet microfluidics with the T-junction capillary system. The various porosities of MOFs were achieved by adding polystyrene particles in the precursor solutions, which were composed of metal salt and organic ligand. The synthesis reaction

was performed with different temperatures in the range of 45-90°C and took 36 h. Conventional methods of MOFs synthesis demand high temperatures, considerable time and also some additional sources of energy (microwave-assisted, ultrasound) to accelerate the synthesis.

Microfluidics enables all the benefits mentioned in other sections describing different types of optically responsive materials (reduced temperature, time and reagent consumption, as well as automation of synthesis route). Furthermore, the droplet microfluidic synthesis approach allows one to obtain the controllable porosity of MOFs since each drop acts as a small reaction chamber, where it is much easier to control important synthesis parameters and concentration of the reagents.

Conclusion on chapter 1

In the 1 st Chapter the main concepts used in the work are considered. Thus, the contemporary state of the material synthesis using microfluidic technology are given with an emphasis on the perovskites and metal-organic frameworks synthesis. Microfluidic synthesis has many specificities including mixing (diffusion) in the continuous flow, as well as interfacial tension affection onto droplets formation in the flow. These and others properties of the microfluidic system for synthesis are discussed.

Additionally, examples of conventional synthesis methods for perovskites and MOFs are given. All the above mentioned conventional methods of synthesis of both perovskites and MOFs operate with macroparameters, which affects the quality of the synthesized materials. When considering the formation of micro/nanomaterials, their morphology (monodispersity, sizes, shapes) and compositional features of each component are becoming increasingly important at the micro/nanoscale. Because of this, precise control of the synthesis conditions of such materials are crucial to obtain the required morphology and physico-chemical properties.

Therefore, a scalable, universal, and reproducible synthesis method for perovskite particles of various sizes, from nanometers to micrometers, remains an issue. Importantly,

the obtained particles should have a narrow size distribution and be dispersed in a tube, which enables their deposition on any arbitrary surface. A promising approach for the controllable synthesis of various nanostructured materials with specific predefined physicochemical properties is microfluidics.

This thesis is devoted to a detailed research of microfluidic approach for the synthesis of high crystalline optically responsive materials.

CHAPTER 2. Fabrication and characterization

In the 2nd chapter a list of the used materials for optically active materials synthesis, the geometry and design of MFCs for the synthesis of optically responsive materials by microfluidics (in a continuous flow regime and droplet microfluidics), as well as a description of the synthesis setup, and detailed protocols for the synthesis of optically responsive materials are given.

Along with that the chapter contains the methods of preparation of synthesized materials for further analysis, the conditions of structural properties analysis (SEM, TEM, pXRD, etc.), and the conditions for the optical properties' analysis (lasing, PLQY, Raman spectroscopy, etc.).

2.1. Materials

2.1.1. Materials for perovskite synthesis

For perovskite synthesis the following reagents were used: Cesium bromide (CsBr, 99.999%, Sigma-Aldrich), lead (II) bromide (PbBr2, 99.999%, Lanhit), anhydrous dimethyl sulfoxide (DMSO, 99.8%, Alfa Aesar), isopropyl alcohol (IPA, technical grade, 95%, Vecton), toluene (technical grade, Vecton), oleylamine (OLA, 95%, Sigma-Aldrich), oleic acid (OA, 90%, technical grade, Lenreaktiv), vaseline oil (VO, technical grade, Lenreaktiv), fluorinated fluid Fluoridrop (Dolomite Microfluidics), emulsion stabiliser FluoSurf (Dolomite Microfluidics).

Preparation of the CsPbBr3 perovskite precursor: 64 g of CsBr was mixed with 110 g of PbBr2 and 3 mL of DMSO in a N2-filled glovebox and shaked for 10 min to obtain a clear solution. The prepared perovskite precursor was further used for the continuous flow synthesis.

Preparation of Cs-oleate: 0.407 g Cs2CO3 and 20 mL of VO were added into 100 mL 2-neck flask and degassed at 120 oC and under magnetic stirring of 390 rpm for 1 h in the

Schlenk line. Further 1.3 mL of OA was injected at 140 oC under N2 and stirred until all cesium salt dissolved completely.

Preparation of lead bromide: 0.07 g of PbBr2 and 7 mL of VO were added into 100 mL 2-neck flask degassed at 120 oC under magnetic stirring of 390 rpm in the Schlenk line. After 1 h, 0.5 mL of OLA and 0.5 mL OA were added under vacuum and degassed for 15 min. Then the temperature was increased up to 140 oC and further stirred under N2 until full dissolving of PbBr2.

Preparation of mixedhalide (leadbromide/iodide): 0.066 g of PbBr2, 0.083 g of Pbl2 and 10 mL of VO were added into 100 mL 2-neck flask degassed at 120 oC under magnetic stirring of 390 rpm in the Schlenk line. After 1 h, 1.6 mL of OLA and 1.6 mL of OA were added under vacuum and degassed for 15 min. Then the temperature was increased up to 140 oC and further stirring continued under N2 until full dissolving of lead salts.

2.1.2. Materials for MOF synthesis

Reagents for MOF synthesis were purchased from Merck: CuSO^5H2O (98%), benzene-1,3,5-tricarboxylic acid or trimesic acid (95%, H3BTC), N,N-Dimethylformamide (99,8%, DMF), ethanol (99,5%, EtOH), acetic acid (96%, CH3COOH).

To perform microfluidic synthesis metal salts and H3BTC were preliminarily dissolved in DMF. For this, 1.67 mmol of CuSO^5H2O (98%) was added into 10 mL of DMF each and ultrasonicated for 15 min. Similarly, 1.13 mmol of H3BTC was added into 10 mL of DMF and also ultrasonicated for 15 min.

For solvothermal synthesis of MOF HKUST-1 powder CuSO^7H2O (1.01 g, 3.5 mmol) was weighed and placed into 20 ml vial. Then 12 ml of deionized water was added, and vial was immersed in an ultrasonic bath for 10 min. Similarly, the solution of H3BTC (0.42 g, 2.0 mmol) in 12 ml of EtOH was prepared. The solution mixture was placed into 40 ml vial and hermetically sealed with a lid with a rubber septum to exclude the

interaction with the external environment and create excess pressure in the vessel. The solution mixture was heated to 1200C and kept for 12 hours. After 12 hours the reaction mixture was cooled down to room temperature. The resulting powder was separated from the mother liquid by filtration, then it was repeatedly washed 5 times with EtOH. The washed powder was dried in the air. A blue homogeneous powder was obtained [84].

For solvothermal synthesis of MOFHKUST-1 single crystal Cu(NO3^3H2O (0.16 g, 0.67 mmol) was dissolved in mixture of 1 ml H2O and 1 ml DMF. Solution of H3BTC (0.08 g, 0.38 mmol) was obtained by dissolving acid powder in 1 ml of EtOH. Then the solution mixture was placed into 20 ml vial. Further, 4 ml of glacial acetic acid was added, the vial was hermetically sealed and heated to 55oC and kept for 72 hours. After 72 hours the reaction mixture was cooled to room temperature and then washed 6 times with EtOH. Obtained crystals were dried in the ambient condition[70].

2.3 Microfluidic chip fabrication

2.3.1. Soft lithography

Microfluidic chips (MFCs) were fabricated by a standard "soft lithography" method using polydimethylsiloxane (PDMS, SYLGARD™ 184 Silicone Elastomer kit, Dow Corning, USA), that consists of two parts: silicone elastomer base (PDMS base) and curing agent. For this, a PDMS mixture of 1:10 ratio of curing agent to PDMS base was mixed and degassed in a vacuum chamber for 10 min. Then the uncured PDMS was poured onto a silicon mold with a photoresist structure and cured at 80 0C during 2 h. Afterwards, the PDMS replica was sealed to a glass substrate (Menzel Microscope Coverslips) by oxygen plasma surface activation and placed on a hot plate with T=90 0C for 15 min.

For the continuous flow synthesis two geometries are investigated: (i) The geometry of an MFC consisted of two inlets for Me salt and an organic ligand, a channel with a widening for better mixing and preventing the clogging due to the crystals deposition, and an outlet (Figure 17).

Figure 17 - MFC for the MOF synthesis via continuous flow. (A) Geometry of the MFC. (B) Digital image of an MFC.

(ii) MFC had a two-channel topology with a thin interlayer (1 mm in length, 50 ^m in thickness), as presented in Figure 18. The configuration of MFCs presented valve design.

Figure 18 - MFC for continuous flow synthesis of perovskites. (A) Geometry of the

MFC. (B) Digital image of an MFC.

(iii) In case when the droplet microfluidics is employed the MFCs preparation consists of an additional step such as surface treatment. This step is crucial for the controlled emulsification.

For further wettability change, an MFC was loaded with either polyvinyl alcohol (in case of water as a continuous phase), or Fluoridrop (in case of fluorinated oil as a

continuous phase) by a syringe. Finally, microchannels of an MFC were cleaned by the air blow-gun.

The geometry of another MFC is presented in Figure 19. and contains three inlets: inlet 1 - for continuous phase (either distilled water or fluorinated oil), inlet 2 - for cesium oleate dispersed in VO, and an outlet, inlet 3 - for lead bromide or mixed halide dispersed in VO.

Figure 19 - (A) Geometry of the MFC. (B) Digital image of an MFC.

2.4. Synthesis setup

2.4.1. Continuous flow

a. MOF in a continuous flow

To perform microfluidic synthesis metal salts and H3BTC preliminarily dissolved in DMF as described in §2.1. Further, prepared solutions of precursors were loaded into different channels of MFC by using syringe pumps (Pump 11 Elite, Harvard Apparatus, USA) connected to an MFC through a capillary system. The outflow stream was connected to the centrifuging tube. The temperature variation was realized by placing an MFC onto a hotplate MR Hei-Tec (Heidolph, Germany). Schematic illustration, as well as a photo of an experimental setup are presented in the Figure 20.

Figure 20 - Synthesis setup. (A) Experimental setup for the synthesis at room temperature. (B) Experimental setup for the synthesis with temperature variations. The schemes include microfluidic chip (1), syringe pumps (2), collection tube (3), optical

microscope (4), and hot plate (5).

b. Perovskite synthesis in a continuous flow

To perform a synthesis experiment, syringe pumps Harvard Pump 11 Plus (Harvard Apparatus, USA) were used. The IPA and perovskite precursor were loaded into different channels of MFC with controllable flow rate (v), while one of the drain holes was blocked, so the flow penetrated through the valve/porous membrane in the interlayer. The products of the reaction were collected in the centrifuge tube filled with 100 ^L of toluene (Figure 21).Thus, a suspension of perovskite particles in toluene was obtained.

Syringe pump

Figure 21 - Schematics of the microfluidic system for synthesis of perovskite particles.

Varying the perovskite precursor rate with a constant IPA rate (IPA to perovskite precursor ratios), perovskite particles with different morphologies were obtained. The flow rates were adjusted experimentally in order to obtain perovskite particles with defined morphologies (Table 2).

Table 2. Samples obtained in the valve configuration of MFC. Variations of IPA to perovskite precursor ratios were used.

V IPA (mm/s) V perovskite precursor (mm/s) Ratio IPA:perovskite precursor

0.29 0.33 1:1.11

0.29 0.37 1:1.25

0.29 0.38 1:1.29

0.29 0.39 1:1.33

0.29 0.41 1:1.38

0.29 0.42 1:1.43

0.29 0.44 1:1.48

0.29 0.45 1:1.54

0.29 0.49 1:1.66

0.29 0.59 1:2

2.4.2. Droplet microfluidics

For the synthesis of perovskite particles by a droplet microfluidics the reagents were supplied by syringe pumps. The droplet generation process was then visualized and controlled via optical microscope. Further, two schemes of synthesis are described.

a. For the synthesis of perovskite nanoparticles

Precursors were supplied into the MFC using syringe pumps (Figure 22). The syringes system consisted of a homemade syringe pump (1) that supplied a continuous phase, and two Harvard Apparatus syringe pumps (2.1, 2.2) that supplied dispersed phase (perovskite precursors). To heat the precursors, homemade heaters for the syringes 2.1 and 2.2 were used. An MFC (4) with a heating substrate (5) was placed under the optical microscope (3). Heating substrate temperature was varied using a power source (6) and controlled by a thermocouple (7). The synthesis product was collected in a centrifuging tube (8).

Figure 22 - Experimental setup for synthesis of perovskites with fluorinated oil as a continuous phase. (A) Experimental setup. (B) MFC placed on a heater with a fixed thermocouple. (C) Schematic representation of the experimental setup.

b. For the synthesis of submicrometricperovskiteparticles

Precursors were supplied into an MFC using syringe pumps (1 on the Figure 23), the temperature of which was maintained with installed heaters for the syringes (1.1) and (1.2). An MFC (2) was placed under the optical microscope (3) and the synthesis outcomes were collected in a centrifuging tube (4).

Figure 23 - Experimental setup for synthesis of perovskites with a distilled water as a carrier fluid. (A) Experimental setup (inset: homemade heater for the syringe). (B) MFC connection to the syringe pumps via capillary system. (C) Schematic representation of

the experimental setup.

2.5. Structural characterization technologies

Further characterization methods applied in this work and sample preparation for each measurement are described in details.

2.5.1. Electron microscopy

Scanning electron microscopy (SEM) measurements were carried out using MERLIN (Carl Zeiss) with an acceleration voltage of 2 kV. Perovskite structures were deposited onto Si substrates for the analysis. Obtained SEM images were analyzed to estimate the average linear size of the perovskites with the ImageJ software.

Transmission electron microscopy (TEM) analysis was performed using the Zeiss Libra 200 FE transmission electron microscope at the accelerating voltage of 200 kV. To prepare a sample, 3 ^L of the obtained perovskite particles was dropped onto a carbon-coated copper grid, and the solvent was removed by evaporation at room temperature.

2.5.2. Powder X-ray diffraction (pXRD)

pXRD measurements were performed (1) on a Shimadzu 7000-maxima X-ray diffractometer with a 2 kW characteristic Cu-Ka (Ka1 X = 1.54059 A) X-ray radiation source and a Bragg-Brentano goniometer geometry. The angular resolution during the analysis was 0.01 degree at a scanning speed of 5 degrees/min.

The second (2) diffractometer used in this study is XRD-7000 Maxima diffractometer, Shimadzu with Bragg-Brentano geometry of measurements. The source of characteristic radiation was Cu K-Alpha, X = 1.5406 nm. The scanning step was 0.02 degree, the exposure time was 1.2 s.

2.6. Optical characterization technologies

2.6.1. Photoluminescence (PL)

PL images of the dried samples (previously deposited onto a glass coverslip) were obtained using an Axio Imager A2m (Carl Zeiss) microscope with 10*, 50*, and 100* objectives (Carl Zeiss EC Epiplan-NEOFLUAR). A mercury lamp (360 nm) was used as an excitation source coupled with a long-pass filter with a cut-on wavelength of 420 nm.

2.6.2. Quantum yield of photoluminescence (QYPL)

To evaluate QYPL, time-resolved PL decay curves for the wavelength range from 0.45 ^m to 1 ^m were recorded using a PicoHarp 300 TCSPC module (PicoQuant). To pump an individual perovskite nanowire (NW) with size 0.38x0.38x5.2 ^m pulsed laser excitation (Xex=430 nm, t=150 fs, f=1 MHz) from Orpheus optical parametric amplifier (Light Conversion) was focused on the sample surface at normal incidence by a long working distance objective (Mitutoyo 50x, NA=0.42) providing the NW with uniform irradiation (Gaussian distribution with FWHM of 10 ^m). For attenuation of excitation light intensity, a pair composed of a Glan prism (GL10-A, Thorlabs) and a half-wave plate (WPMH10M-532, Thorlabs) was exploited. To determine the average power of the excitation beam, a detector (PD300-TP, Ophir) with a data processing controller (NOVA II, Ophir) was employed. PL signal was collected by the same objective, separated from residual excitation light using a longpass filter (FELH0450, Thorlabs), and sent towards a single-photon avalanche diode (Micro Photon Devices) possessing photon timing resolution of 50 ps (FWHM).

Further, to evaluate QYPL we approximated the experimental curves by equations that are solutions for the ABC model proposed by Y. Shen et al. [6]:

C& = -An - Bn2 - Cn3 (12)

where n is a charge carrier density, A, B, and C are trap-assisted, bimolecular, and Auger recombination rate constants, respectively. At low fluences, bimolecular and Auger recombination do not contribute to charge carrier dynamics (B = 0, C = 0), and, thus, the normalized PL decay curve has monomolecular behavior and can be approximated by the following equation:

^ = exp(-At) + const (13) n0

where no is the initial charge carrier density. Increase in pumping power (up to 60 times) results in remarkable acceleration of the PL decay owing to the contribution of bimolecular recombination to the charge carrier dynamics (B > 0). However, this excitation pumping power still are not high enough to make considerable the Auger recombination (C = 0). The solution for (1) in this case is given by:

In the meanwhile, photoluminescence quantum yield can be calculated from the

bimolecular PL decay curve at certain pumping power.

To directly measure QYPL spectrofluorophotometer Shimadzu RF-6000 with an integrating sphere unit was used. The measurements were taken with excitation wavelength X = 365 nm, data interval 1, scan speed 6000 nm/min, excitation and emission bandwidths are 5 nm.

2.6.3. Lasing

Stimulated emission spectra was measured from a single CsPbBr3 structure at room temperature using experimental setup, shown in Figure 24. A femtosecond (fs) laser (PHAROS, Light Conversion) with repetition rate 10 kHz and pulse duration of 220 fs, coupled with a broad-bandwidth optical parametric amplifier (Orpheus-F, Light Conversion) was used as excitation source. The laser beam was focused on a sample surface in the around 10 ^m diameter spot with 10x microscope objective (Mitutoyo Plan APO NIR, NA = 0.26), whereas the signal was collected from the top view with 50* objective (Mitutoyo Plan APO VIS, NA = 0.55) and sent to the Andor Kymera 328i spectrograph coupled with the Andor iDus CCD camera. For one-photon lasing excitation wavelength was set to 525 nm, while for two-photon varied in the 900-980 nm range. For attenuation of excitation light intensity, a pair composed of a Glan prism and a half-wave plate was exploited. A 600 grooves/mm grating and a pinhole with 100 ^m diameter, adjusted to the spectrograph, were used to study the laser emission.

expression QYPL =

n0B

where n0B the value is derived from the approximation of the

;+n0B'

Figure 24 - Optical scheme for lasing measurements. Designations: LP800 - long pass filter 800 nm; HWP - half-wave plate; GP - Glan prism; BS1/BS2 - 50:50 Beam splitter; Obj x10 - Objective x10, NA=0.26; Obj x50 - Objective, x50 NA= 0.55; M -mirror; SP600 - short pass filter 600 nm; L - lens to spectrograph; red and green stripes

represent pump and signal paths, respectively.

2.6.4. Raman-spectroscopy

To measure Raman scattering, a coherent light from HeNe laser (632.8 nm) was focused on single crystals of CuBTC-1M OF using a high-aperture objective (Mitutoyo 100x 0.9 N.A.). The scattering signal has been collected then by the same objective and transferred to HORIBA LabRAM confocal spectrometer with a water-cooling chargecoupled device (CCD, Andor DU 420 AOE 325) with a 600 g/mm diffraction grating (Figure 25).

Figure 25 - Optical scheme of Raman spectroscopy measurements

The Raman mapping measurements were performed using the motorized stage (Thorlabs) with 1 ^m step and 100 s signal counting for each spatial position. The spatial plane cut for each peak of the Raman spectrum that corresponds to chemical bonds are presented in the §3.

2.7. Dye loading

First, 77 mg of bovine serum albumin (BSA) were mixed with 20 ml of phosphate-buffered saline (PBS), separately 5 mg of Cy5 were dissolved in 2.2 ml of DMSO. Then, two solutions were mixed at 4°C for 24 h. Further, the required volume of the initial Cy5-BSA solution was dissolved in water.

Afterwards, 0.27 mg of MOF from microfluidic synthesis and solvothermal powder synthesis were added into each sample with dye and kept on a shaker Heidolph Vibramax 100 (Heidolph Instruments, Germany) for 1 h. For further measurements of absorbance, incubated samples were centrifuged at 10 000 rpm for 3 minutes using Sigma 1-14 mini centrifuge (Sigma Laborzentrifugen GmbH, Germany).

Dye adsorption was determined as the difference between its initial and equilibrium

concentrations in a water solution after mixing with a sorbent from the equation:

n _ (c0-cads)v n

where Qe is the magnitude of Cy5-BSA adsorption, Co and Cads are the initial and equilibrium concentrations of Cy5-BSA in a water solution, V is the volume of the solution, and m is the mass of the MOF. C0 and Cads were calculated using Beer-Lambert law:

A = £Cads I (16)

where £ is molar absorption coefficient, I is an optical path length. Measurements of A were performed using the UV-Vis-NIR spectrophotometer Shimadzu UV-3600 Plus (Shimadzu, Japan).

2.8. Cells study

Part of the research includes the study of the effect of synthesized materials on the cells properties. The methods and types of cells that were used in the work are described below.

2.8.1. Cells dye loading

For in vitro experiments: Alpha Minimum Essential Medium (Alpha-MEM) and phosphate-buffered saline (PBS) were obtained from Lonza, Switzerland. AlamarBlue cell viability reagent was purchased from Invitrogen, USA. Rhodamine B (RhB, >95%) were purchased from Sigma-Aldrich.

For cell experiments, murine melanoma cells (B16-F10) were used. The cells were cultivated using AlphaMEM culture medium supplemented with 10% of vol. fetal bovine serum and 2 mM UltraGlutamine I at 37°C. The cell line was maintained in an incubator containing 95% air and 5% CO2.

2.8.2 Toxicity study

The toxicity of synthesized MOFs (HKUST-1) was estimated using the AlamarBlue assay. For this purpose, B16-F10 cells were seeded into a 96-well plate (V = 200 ^L, 100 000 cells/well). On the next day, MOFs were added to the cells at different

MOF:cell ratios. After 24 hours, the cells were washed twice with PBS to remove non-internalized MOFs and fresh cell culture medium supplemented with 10% vol. of AlamarBlue was added. After 4 h of incubation, the medium containing AlamarBlue was analyzed by measuring the absorption spectra at 570 and 600 nm with UV-vis spectrophotometer (Thermo Scientific Multiskan GO). Experiment was repeated three times to obtain the mean value and the standard deviation.

2.8.3. Uptake study

To visualize the internalization of MOFs (HKUST -1) by cells, B16-F10 cells were seeded into a confocal cell imaging dishe (Eppendorf, d = 35 mm) at the amount of 500 000 cells per dish and left overnight. Next day, MOFs loaded with Cy5 were added to the cells at the ratio 5 crystals per cell. After 24 h of incubation, cells were washed twice with PBS and fixed with 4% paraformaldehyde. Then cells were stained using Rhodamine B. For that, 10 microliters of RhodamineB were added to each dish with PBS and cells. The images of the stained cells with internalized MOFs were obtained using CLSM (Carl Zeiss LSM 710) in Z-stack modus. The stained cells were visualized using an argon laser at 514 nm (emission filter BP 520-570 nm), and MOFs were imaged using an argon laser at 633 nm (emission filter BP 640-700 nm).

10 nm

Figure 26. CLSM images of B16-F10 cells incubated with MOFs for 24 h. B16-F10 cells were stained with RhB (red) and HKUST-1 were fluorescently labeled with Cy5-BSA (green).

Conclusion on chapter 2

Chapter 2 outlines techniques for examining the structure and optical properties of the samples obtained, as well as providing various methods for further analysis of their characteristics. The preparation of samples for these methods is also discussed in details. The complexity and relevance of these techniques reinforce the validity of the data presented in the dissertation.

CHAPTER 3. Continuous flow synthesis

In this chapter a comprehensive investigation of the continuous flow synthesis approach for the controlled synthesis of optically responsive crystals, along with a detailed characterization of the resulting optically responsive crystals using methods of structural analysis are given. Furthermore, an in-depth examination of the optical properties of the synthesized crystals is presented.

The description of the unique features of continuous flow synthesis and comparison to segmented and droplet flow are outlined in Chapter 1 of the thesis.

3.1 Continuous flow synthesis of MOF

For the synthesis of MOF HKUST-1 by the continuous flow method in an MFC solutions of metal salt and organic ligand were prepared in advance as described in §2.1.2. Further, the obtained solutions were loaded into glass syringes and supplied to the MFC (Figure 17) via polytetrafluoroethylene (PTFE) tubing and syringe pumps as described in §2.4.1. Additionally, solvothermal synthesis of MOF HKUST-1 was performed as described in §2.1.2.

3.1.1. Product yield estimation

Further, the product yield was estimated to compare the effectiveness of microfluidic and solvothermal methods of synthesis. The product yield was calculated based on the used amount of Me salt by the comparison of the theoretically calculated product yield and the practically obtained.

mexperimantal . WQ% (17)

^theoretical

Obtained results are given in Table 3.

Table 3. Product yield of MOFs synthesis for microfluidic and solvothermal methods.

Synthesis method 0oC 22oC 50oC 70oC 90oC 110oC

Microfluidic 35% 54% 32% 32% 30% 41%

Solvothermal 48% powder (120oC) and 37% crystal (55oC)

The results indicate that the microfluidic synthesis method yields a higher product yield at room temperature (T = 22oC) compared to the solvothermal synthesis method. However, at other synthesis temperatures using the microfluidic method, the product yields are lower than those obtained through solvothermal synthesis.

3.1.2. Mixing in the channel of MFC

Numerical simulation of the mixing process of MOF precursors in the channel of an MFC demonstrated that the widening of the channel plays the role of a passive mixer. Due to this, the complete mixing of precursors is achieved already at the region of the channel's widening (Figure 26). The fulfilling of the MFC channel occurs in 40 seconds. Taking into account the length of the output capillary and the constant flow rate of reagents, the synthesis time equal to 30 min is determined.

Figure 26 - Numerical simulation of reagents mixing in the MFC channel during the first 40 seconds after filling the channel

with

reagents.

To study the possibilities of the synthesis reaction control, the reaction temperature was varied. For these purposes, the MFC was placed on a heating substrate (Figure 20) or on a cooling platform to create a temperature equal to T = 0oC.

3.1.3. Structural characterization of MOF HKUST-1

The obtained structures of the HKUST-1 were characterized by scanning electron microscopy (SEM) (Figure 27).

As can be seen from the figure, the resulting structures have a characteristic octahedron shape. With an increase in the synthesis temperature, aggregation is observed, which is explained by an increase in crystallization centers (crystal nuclei) and accelerates the crystal growth reaction.

Figure 27 - Structural characterization of the obtained HKUST-1 at the temperature

variation

An increase of crystallization centers under conditions of enhanced mixing due to the channel widening, leads to rapid crystal growth. Thus, the formed nuclei aggregate and form agglomerates, that are shown in Figure 27.

An estimation of the average linear size of the agglomerates shows that with increasing temperature, the linear size of the agglomerate increases as well.

Figure 28 - Geometrical size of the MOF HKUST-1 dependency on the temperature of

the synthesis fitting.

Further, for in-depth analysis of the synthesized with temperature variation HKUST-1 structure, the MOFs were analyzed by powder X-ray diffraction (pXRD). The obtained diffractograms indicate a change in the quality of the structure when applying different temperatures (Figure 29). Thus, the diffractogram of the sample at T = 0oC is characterized by a low intensity of reflexes and noticeably wider reflexes. When the temperature increases to T = 22oC, which corresponds to room temperature, the reflexes are pronounced and fully correspond to the reference values. Diffractograms of structures obtained at higher temperatures equal to 50-90 oC indicates a high crystallinity of the samples. An increase in the synthesis temperature to 110oC probably led to the amorphization of the reaction product.

Figure 29 - Powder XRD of the obtained MOF HKUST-1 synthesized at different

temperatures.

3.1.4. HKUST-1 defects study

To study closer the defects formation in a single crystal of HKUST-1 the Raman-mapping was performed. The Raman mapping measurements were performed using the motorized stage (Thorlabs) with 1 ^m step and 100 s signal counting for each spatial position.

Figure 30 - Raman-spectroscopy characterization of the obtained MOFs. (A) Raman spectrum of HKUST-1 (inset: optical image of a HKUST -1 single microcrystal); (B) Raman mapping of a single crystal of HKUST -1.

A microscale single crystal was mapped by confocal Raman microscopy with a spatial step of 1 ^m within the region of 100-2000 cm-1. By addressing the Raman modes

5323233002322348484802025323232323232323484800020202535323232323232348020202020253532353234800000202020202

to specific vibrations of HKUST-1 (Table 4) and taking into account the wide spectrum of its defects (Table 5), the following assumptions about the growth of the MOFs inside the MFC were made.

Table 4. Chemical assignments of vibrational modes for the Raman spectra of HKUST-1.

Peak position, cm-1 Peak Assignment Ref.

173 v(Cu-Cu) Ref.[79; 111]

266 v(Cu-Ow)

495 v(Cu-O)

730,814 Y(C-H)

929 v(C-C)

997,1610 v(C=C)

1430,1580 v(COO)

v - stretching vibration; y - out-of-plane bending; Ow - oxygen of H2O

The area in the lower right corner of the crystal (highlighted by a white arrow in Figure 31) is characterized by a relatively high concentration of coordination bond Cu-O, Cu-Ow (Table 4) and the modes of the ligand itself, which allows us to speculate about local high-quality crystal seed.

Figure 31 - Raman spectroscopy of a single crystal of MOF HKUST-1 obtained via solvothermal method. (A) Raman spectrum (inset: optical image of single crystal); (B)

Images of a single crystal Raman-mapping.

Further growth of the crystal upwards may be due to the appearance of point defects in the lower region (indicated by the purple arrow) as a metal node reduction and formation of a Cu2+ cluster, which is characterized by a high Cu-Cu signal with a parallel decrease in the intensity of the ligand modes. Such defects have been also observed along the entire perimeter of the crystal (blue arrow). Moreover, we have detected the edge region of the low ligand signal (green arrow), which can be interpreted as a ligand vacancy due to the non-zero Cu-O signal. Intriguing is that the inner region of the crystal

demonstrates an underestimated intensity of the Cu-Cu, Cu-O, and Cu-Ow signals (red arrow), while the intensity of the ligand modes is rather uniform within the volume. This observation possibly indicates a lower concentration of water inside the crystal, as well as a weak degree of coordination due to missing Cu atoms. In the latter case, we exclude the dislocation effect [11; 77; 95], since it would lead to an increase in the C-O-O and Cu-Ow intensities. As a result, the growth process in the chip is accompanied by the formation of a whole spectrum of defects of HKUST-1, which, on the one hand, can slow down the further growth of the crystal, and, on the other hand, serve as active centers [135].

Table 5. HKUST-1 defects.

Defect Details Bonds Ref.

Linear (dislocations) plane dislocations with free COOH groups Breaking coordination bond Cu-OOC in {111} plane Ref.[38]

Linear (dislocations) dislocation growth spirals at {111} facets Breaking coordination bond Cu-OOC Ref.[29; 61]

Point (vacancy) partial metal node reduction (Cu+/Cu2+ dimer defects) due to linker vacancies decreased number of coordination Cu-OOC bonds and increase of Cu-H2O bonds Ref.[35; 60; 62; 63; 78]

Point (vacancy) missing paddlewheel clusters decreased number of Cu-Cu, Cu-Ow and coordination Cu-OOC bonds Ref.[15; 60]

Point (vacancy) Cu2+ cluster decreased number of Cu-H2O bonds and increased Cu-Cu bind Ref.[111; 124]

Surface defect missing carboxylates on the external surfaces Decreasing Cu-OOC bond Ref.[46, c. 1]

Surface defect Preferential surface {111} Less broken coordination Cu-O bonds in {111} Ref.[3; 120]

Disorder defective Cu sites aggregate in a stable supramolecular H-bonding structure Cu-O replaced by Cu-H Ref.[93]

Disorder gelation All bonds are preserved with broadening peaks Ref.[43; 75]

In contrast to such defective MOF crystals, reference HKUST-1 crystals obtained by slow solvothermal synthesis are characterized by a lower concentration of surface defects (Figure 32). A uniform distribution of the bonds corresponding to organic part and coordination interaction over the crystal surface cannot, on the one hand, indicate the complete absence of point defects; on the other hand, this indicates significantly lower distortions and inhomogeneities structures that are observed for HKUST-1 crystals obtained by microfluidic synthesis. This statement we than confirm by the difference in the efficiency of dye sorption on the MOF surface (Figure 33).

0 250 500 750 1000 1250 1500 1750

Raman Shift (cm'1)

Figure 32 - Optical characterization of HKUST-1 MOF synthesized by solvothermal synthesis: (A) Raman spectrum of HKUST-1 (inset: optical image of HKUST-1 single crystal); (B) Images of a single HKUST-1 crystal Raman-mapping.

3.1.5. Defects of HKUST-1 for dye adsorption

Further, the adsorption properties (Qe) of MOFs were investigated. To compare dye adsorption capacities, HKUST-1 synthesized by solvothermal method alongside with microfluidically synthesized ones were used. For loading, a physical adsorption approach was used. This method is widely used for loading of different biomolecules. It is worth noting that knowing the adsorption properties is essential in case of further application of MOFs as drug delivery carriers [21; 96; 132]. High molecular weight Cyanine 5 (Cy5)

conjugated with bovine serum albumin (BSA) and low molecular weight Rhodamine B (RhB) were used as model loading molecules.

Figure 33 - Cargo loading into MOFs and cell studies. (A) Nonlinear fit of adsorption isotherm curve for Cy5-BSA adsorption of MOFs obtained via microfluidic method of

synthesis (1) and via solvothermal method of synthesis (2); (B) Cells viability when incubated with Cy5-BSA loaded HKUST-1 at various concentrations; (C) Representative confocal laser scanning microscopy (CLSM) image of a single HKUST-1 crystal labeled with Cy5-BSA.

Confocal laser scanning microscopy (CLSM) images of fluorescently labeled MOFs confirm successful loading of dye molecules (Figure 33C). CLSM image of a loaded HKUST-1 microcrystal indicates the surface loading of MOFs with fluorescence dye. These results are in good agreement with the Raman mapping of defects on the

surface and also correlate qualitatively with the fact that the size of the molecules is large enough for potential penetration into the pores of MOF with a large number of defects. For the estimation of the adsorption efficiency of the synthesized MOFs, particles were then incubated in the Cy5-BSA solution at different concentrations (0.9 ^g/mL - 11.2 ^g/mL) for 1 h. Adsorption isotherms were then plotted with the Langmuir fitting represented by the following equation [87]:

/-> _ QmaxKCads /1 o\

1+bC'ds ( )

where b is the Langmuir isotherm constant, Qmax is the theoretical monolayer saturation capacity of the Cy5-BSA, and Cads is the equilibrium concentration of Cy5-BSA. A detailed description of the calculation of the adsorption capacity is given in §2.7. According to the obtained Langmuir isotherm, it can be seen that high amount of Cy5-BSA can be adsorbed onto the surface of MOFs and the loading capacity of MOFs is increased followed by loading saturation of MOFs (Figure 33).

The comparison reveals that the Qe for the MOFs obtained via microfluidic method of synthesis is higher than for the MOFs obtained via solvothermal method of synthesis by approximately 2 times. Adsorption of Cy5-BSA onto HKUST-1 can occur due to two factors: (i) electrostatic interaction between charged loading molecules and surface of MOFs and (ii) capture on defects with uncompensated charges [10; 128].

Further, toxicity of HKUST-1 loaded with Cy5-BSA was evaluated on murine melanoma cells (B16-F10) using AlmarBlue assay. For this, MOFs were added to cells at different concentrations of MOFs (0:1-25:1 MOFs/cell) and incubated for 24 h. There was almost no cytotoxicity detected at the lowest concentration of added MOFs (1 microcrystal per cell). With the increase of MOFs concentration, the viability of cells was slightly reduced up to 80-75%, but it is still appropriate for these concentration ranges. Thus, the obtained data reveal non-toxicity of HKUST-1 at added concentrations, which is in agreement with the previously published works [80]. In order to visualize cell-MOFs interaction, CLSM was applied. For this, cell membranes were stained with Rhodamine B (RhB) and HKUST-1 loaded with Cy5-BSA were incubated with cells for 24 h. After

incubation, the cells with associated MOFs were analyzed with CLSM using the Z-stack option to visualize MOF crystals uptake by cells (Figure 34).

Figure 34 - Representative CLSM images of cells incubated with Cy5-BSA loaded CuBTC-1.

CLSM images revealed the stability of the obtained HKUST-1 loaded with Cy5-BSA in cell culture medium. Three-dimensional CLSM image (Figure 34) of a scanned cell with internalized HKUST-1 provides visualization of a cell-HKUST-1 interaction in the x-y section at given z-location.