Synthesis, Assembly, and Applications of Hybrid Nanostructures for

Review Cite This: Chem. Rev. 2017, 117, 12942-13038

pubs.acs.org/CR

Synthesis, Assembly, and Applications of Hybrid Nanostructures for Biosensing Shuaidi Zhang, Ren Geryak, Jeffrey Geldmeier, Sunghan Kim, and Vladimir V. Tsukruk* School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332-0245, United States ABSTRACT: The robust, sensitive, and selective detection of targeted biomolecules in their native environment by prospective nanostructures holds much promise for realtime, accurate, and high throughput biosensing. However, in order to be competitive, current biosensor nanotechnologies need significant improvements, especially in specificity, integration, throughput rate, and long-term stability in complex bioenvironments. Advancing biosensing nanotechnologies in chemically “noisy” bioenvironments require careful engineering of nanoscale components that are highly sensitive, biorecognition ligands that are capable of exquisite selective binding, and seamless integration at a level current devices have yet to achieve. This review summarizes recent advances in the synthesis, assembly, and applications of nanoengineered reporting and transducing components critical for efficient biosensing. First, major classes of nanostructured components, both inorganic reporters and organic transducers, are discussed in the context of the synthetic control of their individual compositions, shapes, and properties. Second, the design of surface functionalities and transducing path, the characterization of interfacial architectures, and the integration of multiple nanoscale components into multifunctional ordered nanostructures are extensively examined. Third, examples of current biosensing structures created from hybrid nanomaterials are reviewed, with a distinct emphasis on the need to tailor nanosensor designs to specific operating environments. Finally, we offer a perspective on the future developments of nanohybrid materials and future nanosensors, outline possible directions to be pursued that may yield breakthrough results, and envision the exciting potential of high-performance nanomaterials that will cause disruptive improvements in the field of biosensing.

CONTENTS 1. Introduction 2. Current Biosensing Technologies and Critical Needs for Improvement 2.1. Specificity and Multiplexing Capability 2.2. Biofouling and Long-Term Stability 2.3. Throughput and Dynamic Range 2.4. General Design Principles of Nanostructured Sensing Elements 3. Nanostructured Reporters 3.1. Noble Metal Nanostructures 3.1.1. Synthesis and Fabrication of Plasmonic Metal Nanostructures 3.1.2. Directed Assembly of Colloidal Nanostructures 3.2. Semiconductor Nanostructures 3.2.1. Silicon Nanowires 3.2.2. 2D Transitional Metal Dichalcogenides 3.2.3. Quantum Dots 3.3. Carbon Nanostructures 3.3.1. Carbon Nanotubes 3.3.2. Graphene-Based Materials 3.3.3. Other Emissive Nanostructures 4. Soft Nanomaterials as Functional Components for Hybrid Nanostructures 4.1. Antibiofouling Components 4.1.1. OEG and PEG-Based Systems © 2017 American Chemical Society

4.1.2. Zwitterionic Films 4.1.3. Polyelectrolyte Systems 4.2. Specificity-Enabling Components 4.2.1. Antibodies and Their Derivatives 4.2.2. Artificial Binding Proteins 4.2.3. Aptamers 4.2.4. Enzymes 5. Integration of Nanostructured Reporters and Transducing Components 5.1. Surface Functionalization Chemistry for Biorecognition Unit Integration 5.1.1. Initial Surface Chemistry of Various Nanostructures 5.1.2. Biofunctionalization through Physisorption 5.1.3. Surface Activation 5.1.4. Direct Covalent Biofunctionalization 5.1.5. Use of Affinity Tags for Biorecognition Unit Incorporation 5.2. Controlling Interfacial Morphologies 5.3. Characterization of Nanoscale Interfacial Morphologies

12943 12944 12946 12947 12948 12949 12951 12951 12952 12955 12957 12957 12958 12959 12960 12961 12961 12962 12962 12962 12964

12964 12965 12966 12966 12967 12967 12968 12970 12970 12970 12971 12972 12973 12975 12976 12979

Special Issue: Bioinspired and Biomimetic Materials Received: February 9, 2017 Published: September 13, 2017 12942

DOI: 10.1021/acs.chemrev.7b00088 Chem. Rev. 2017, 117, 12942−13038

Chemical Reviews 5.4. Materials for the Incorporation and Encapsulation of Biomacromolecules 5.4.1. LbL Microcapsules and Shells 5.4.2. Spray Fabrication of Microcapsules 5.4.3. Interfacial Gelation of Microcapsules 5.4.4. Molecularly Imprinted Polymer Matrices 6. Examples of Nanostructured Biomolecular Detectors 6.1. Electrical-Based Detection with Hybrid Nanostructures 6.1.1. Electrochemical Biosensing with Integrated Hybrid Nanostructures 6.1.2. Biosensors with Engineered Nanopore Membranes 6.1.3. Field Effect Transistor-Based Biosensors 6.2. Optical Detection of Biological Targets 6.2.1. Biodetection with Surface Plasmon Resonances 6.2.2. Localized Plasmon Resonance-Based Biosensors 6.2.3. Biosensing with Surface-Enhanced Raman Scattering 6.2.4. Biosensors Based on Quantum Dot Photoluminescence 6.3. Biodetection through Nanomechanical Transduction 6.3.1. Biosensors Based on Scanning Probe Microscopy 6.3.2. Biofunctionalized Microcantilever-Based Biosensors 7. Nanostructured Labels for Cellular and in Vivo Imaging 7.1. Bioimaging with SERS Tags 7.2. Bioimaging with QD Tags 8. Current Trends and Perspectives 8.1. Summary of Current Achievements 8.2. Perspective and Future Trends Author Information Corresponding Author ORCID Notes Biographies Acknowledgments References

Review

elements. Such performance is possible in part by transitioning to the use of functionalized nanoscale components, which are comparable in dimension to biological molecules and allow for highly compact designs of sensor elements and tailored direct molecule-to-molecule interactions and signal transduction. The advantages of smaller sensing structures with nanoscale dimensions also extend beyond simple scaling considerations. Nanomaterials can be engineered to exhibit distinctive properties that are not present in traditional bulk materials of the same composition. When engineered properly, these unique properties and functionalities can allow highly selective binding events to be detected with nanoscale sensing elements. In addition, the discrete nature of molecular recognition interactions at the nanoscale coupled with the small sizes of the associated nanoscale components can facilitate an unprecedented degree of sensitivity, even single molecule detection can sometimes be realized in near-practical conditions.6 However, past advances in the synthesis, assembly, and integration of functional nanoengineered components into actual nanosensors have not resulted in the much-anticipated practical achievements. Thus, a systematic review on the recent significant results and a summary of current challenges in the field is a timely endeavor. It is helpful here to draw a general comparison to the field of nanomedicine, in which nanoengineered components are widely used either as carriers for targeted drug delivery or as direct therapeutic agents for the physical destruction of malignant tissues (Scheme 1).7,8 Similar to the field of nanobiosensing, nanomedicine-related research has promised dramatic improvements to current medical practices.9,10 After decades of research, some success has been achieved in this field. To date, a few categories of nanocarriers such as liposome-encapsulated doxorubicin and protein-bound paclitaxel have been approved for clinical use (Scheme 1).11−13 However, the transformative innovation promised by selfguided precise nanotherapeutics is yet to happen. In a review article recently published by Chan et al., a comprehensive survey of literature from the past decade has revealed one of the major problems in the field: only very minute fraction (0.7%, median) of nanoparticles have actually reached their intended target.14 This low delivery efficiency was largely attributed to insufficient understanding of nanoparticle pharmacokinetics. Crucial concepts such as enhanced permeability and retention (EPR) effect in tumors are still being debated.15 To date, many fine details of nanoparticle extravasation, macrophage uptake, and clearance remain unclear.14 Overall, we suggest that this incomplete understanding can be traced further back to the lack of in-depth knowledge regarding nanoparticle’s surface chemistry, interparticle interactions, and their interactions with biological elements such as serum proteins from the bioenvironment.14,16,17 We suggest that the emerging field of nanobiosensing can be hampered by the same general incomplete understanding of the complex interfacial nanochemistry. Therefore, further optimization of sensing elements and sensing paths is directly related to devising ways to efficiently tailor nanostructure−biological interfaces. Although, we suggest that the field of exosomatic nanobiosensing will be easier to advance as their application does not to involve active immune responses from hosts. In the end, a complete and systematic understanding of behavior of hybrid nanostructures in complex biological environment will likely be key in overcoming the current obstacles in achieving

12982 12983 12984 12984 12985 12986 12986 12986 12987 12988 12989 12989 12990 12991 12994 12995 12995 12996 12997 12997 13000 13001 13002 13003 13008 13008 13008 13008 13008 13009 13009

1. INTRODUCTION Currently, there are a number of successful, well-established biomolecular detection techniques based on various analytical approaches such as immunoassays, high performance liquid chromatography (HPLC), and mass spectroscopy (MS) that have been extensively developed for rapid, accurate, and selective detection. However, practical limitations for real-life implementation such as cost, mobility, reproducibility, and throughput often lead to difficult decisions on the trade-offs between different aspects of device performance if higher selectivity and further miniaturization are desired for in-field, real-time, and wearable applications.1−5 With the advent of nanoengineered biosensor components and materials and the introduction of new, more sophisticated soft-matter fabrication and directed assembly techniques, prospective novel micro- and nanodevices can be introduced to bypass several of these tradeoffs by utilizing highly sensitive and selective biocompatible 12943

DOI: 10.1021/acs.chemrev.7b00088 Chem. Rev. 2017, 117, 12942−13038

Chemical Reviews

Review

duction shells made of various soft synthetic and biological nanomaterials. To this end, we focus on three interconnected aspects of this integration: the functionalization chemistry and soft-assembly approaches used for different nanostructured surfaces, the design of surface morphologies and material architectures based on those approaches, and the integration of multiple nanostructures into higher-order multifunctional entities. Sections 6 and 7 include examples of current and emerging applications of hybrid nanostructured biosensing elements for different applications. We discuss nanoengineered biosensors that could potentially be implemented in practical devices such as biomolecular sensors for specific detection and as contrast tags for cellular and in vivo imaging. We mostly focus on current results and the use of nanoengineered components under real physiological conditions such as in processed body fluids or in living organisms. Finally, in section 8, we summarize some important findings and briefly present our thoughts and suggestions on critical issues, recent achievements, current directions in target applications, and emerging trends in the field of nanoengineered biodetection as can be concluded from and supported by the results summarized in this review. In conclusion, we identify several directions that will be intensively developed for future generations of commercially competitive nanoengineered biosensors.

widespread commercial success for both nanobiosensing and nanotherapeutics.18 In this review, we collect, discuss, and summarize recent significant results on a wide range of nanoengineered elements relevant to modern biosensing approaches including basic sensing elements, reporters responsible for indicating targeted biomolecular events, signal transducers responsible for selective interactions with the bioenvironment, and auxiliary elements capable of mediating secondary events such as fouling. These components will be discussed in detail including their synthesis, fabrication, assembly, and integration for prospective biosensing applications. We will also briefly review the fundamental principles and mechanisms behind the synthesis of the major elements used in nanobiosensing, their chemical and physical properties, and the corresponding physical phenomena involved in the operation of the nanostructured biosensing elements. Several related examples of current and emerging practical applications in this field will be presented as well, and general trends and perspectives will be discussed and suggested in the final section. In the introductory sections, we also refer to important reviews in respective fields that were published in recent years, including those concerning plasmonic nanostructures,19−30 ligand surface chemistry,31−35 multifunctional hierarchical nanostructures,36−42 and sensor−bio interfaces.43−46 Instead of focusing on one or several important aspects of nanosensor fabrication or a specific type of nanosensor as these reviews have already adequately summarized, we will focus on global aspects: (1) providing a holistic view on the integration of various nanoengineered components, (2) highlighting the most recent advances in synthesis, fabrication, and characterization, (3) summarizing recent integration approaches for reportertransducer elements, and (4) outlining current challenges and future trend for the development of nanobiosensing devices. Overall, our discussion provides updates and comprehensive references across different biosensing-related areas that could benefit readers in various complementary research areas. The review is organized into eight major sections besides this short introductory section. In section 2, we begin by giving a brief overview on current commercially available biomolecular detection methods and provide a rational for the future improvements of biosensing technologies. We then elaborate on the main limitations and competitive disadvantages of current biosensor designs and highlight how developments in nanotechnology and nanomaterials might help affinity-based nanobiosensing overcome these limitations to become mainstream molecular detection methods. In section 3, we introduce a variety of different reporting nanostructured elements based on noble metals, semiconductors, and carbon nanostructures and discuss their relevant optical and transport properties. We place particular emphasis on the assembly of these nanostructures with controlled geometrical organization and physicochemical properties. In section 4, we discuss two different categories of soft organic components for the functionalization of various reporting inorganic nanostructures. We first discuss organic ligands mostly exploited as interfacial passivating agents and then present various biorecognition units and related approaches for their controlled immobilization and selective interactions with complex biological environments. In section 5, we address the critical issues of the controlled integration of nanostructured reporters and organic trans-

2. CURRENT BIOSENSING TECHNOLOGIES AND CRITICAL NEEDS FOR IMPROVEMENT Currently, the majority of biomolecular detection routines adopted for practical applications such as clinical diagnosis, environmental monitoring, and drug discovery are based on “wet” biochemistry assays such as enzyme- and immunoassays.47−49 Biosensing assays readily available on the market today typically operate through incubation with analyte solutions and calorimetrically or spectroscopically measuring the enzyme activity.50 For example, the prevalent enzyme-linked immunosorbent assay (ELISA) technique operates by selectively capturing analytes onto antibodies immobilized in microplate wells and monitoring the subsequent color changes that result from enzymatic amplification reactions.51 Such techniques have proven to be extremely versatile and have been widely used in the clinical detection of viral antibodies.52,53 For quantitative results, chemical assays are often combined with fluorescent, chemiluminescent, or radioactive labels so that the analyte concentration can be quantitatively monitored through the precalibrated label’s photoluminescence or radioactivity.54,55 Another widely used technique, electro-chemiluminescence immunoassays (ECLIA), use two antibodies to sandwich the analyte in solution.56 In this case, one antibody is labeled with an electro-chemiluminescent label, such as a ruthenium complex, while the other antibody is attached to a magnetic microparticle. Eventually the entire sandwich assembly is magnetically collected on an electrode, introduced to a regenerating agent, and electrically excited to luminesce with the photon counts directly corresponding to the concentration of analytes present in the original solution.57 The chemical analysis techniques described above, while effective, have several practical limitations. Their need for reagents and complex lab equipment for analysis provide no portability for point-of-care diagnosis and continuous on-site environmental monitoring.57 The requirements for preassay 12944

DOI: 10.1021/acs.chemrev.7b00088 Chem. Rev. 2017, 117, 12942−13038

Chemical Reviews

Review

straightforward signal transduction process. The reagentless nature of biosensors and their ability to be configured for electrical readouts also make them extremely attractive for realtime detections and miniaturized applications such as wearable sensors. For these reasons, it was widely believed that biosensor technologies would eventually replace biochemical assays to become the mainstream analysis tool for clinical diagnosis and drug discovery and health, environmental, industrial, and agricultural monitoring.60 For example, blood glucose biosensors based on the selective catalysis of glucose oxidation by immobilized glucose oxidase and the amperometric detection of the reaction product (hydrogen peroxide) by a platinum anode have been successfully created. This detection scheme is quantifiable due to the amount of charge transferred being directly proportional to the concentration of H2O2 and concordantly, glucose. Due to their ease-of-use and fast response, blood glucose biosensors have long been established and serve as a vital tool in the management of diabetes.61 Lactate, and urea biosensors based on similar oxidase based electrochemical principles have also seen various degrees of commercial success in point-of-care monitoring of their biomolecular targets.62−67 However, despite abundant research activity in the field and long-term development, the envisioned widespread adoption of versatile biosensor technologies beyond the few current examples has not begun. Except in niche fields, facile biosensor applications have failed to gain prevalent acceptance in the broader sensing community.68 There are several reasons for why progress has stalled: biosensors still lack competitive advantages against traditional analytical methods in terms of versatility, reliability, stability, ease-of-use, and, most importantly, cost and throughput.94 Frequently, practical applications in clinical or pharmaceutical laboratories require reliable detection of various types of analytes with high throughput, which is a significant challenge for traditional biosensors. In Table 1, we collected a few representative examples of current commercially available point-of-care hand-held biosensors for clinical blood analysis on the U.S. market. These biosensors already show impressive performances, being able to detect small molecule target at physiologically relevant levels from undiluted blood within minutes or even seconds. However, as mentioned above, today’s point-of-care biosensors are largely limited to the detection of simple molecular analytes that have relatively high expected concentration in the blood. Also, an important consideration is that the market for the specific analyte has to be large enough to offset the relatively expensive development cost by economies

Scheme 1. Examples of Commercial Nanomedicine. Adapted with permission from ref 13. Copyright 2014 American Chemical Society.

purification and postassay calibration consequently entail that detection cannot be conducted in real-time with high throughput. The use of highly specialized labels makes precise quantification inefficient and relatively costly as well. The presence of labels may also interfere with the molecular binding process, thereby making the binding kinetics difficult to control.58 On the other hand, simple and fast colorimetricbased systems such as ELISA have limited accuracy and sensitivity in analyte quantification.59 There is little doubt that biosensor technologies based on the direct combination of biorecognition units and reporter elements that are capable of on-spot transduction of molecular binding events into easily detectable signals are far more efficient methods to utilize biorecognition units compared with traditional chemical assays (Scheme 2).94 In principle, their accuracy and sensitivity are increased as well due to their

Scheme 2. Recent and Current Biomolecular Detection Approaches Such As Immunoassays, Traditional Biosensors, and Nanobiosensors Composed of Nanostructured Reporters and Functional Transduction Shells

12945

DOI: 10.1021/acs.chemrev.7b00088 Chem. Rev. 2017, 117, 12942−13038

Chemical Reviews

Review

Table 1. Representative Examples of Current Handheld Biosensors company

device

analyte

measurement range

response time

Roche Diagnostics, Switzerland

StatStrip Glucose Hospital Connectivity Meter StatStrip Lactate Connectivity Meter StatSensor Creatinine meter Accutrend Plus system

Abbott Laboratories, USA

i-STAT system

OneTouch Verio IQ meter

0.6−33.3 mM 0.3−20 mM 27−1056 μM 1.1−33.3 mM 0.8−22 mM 3.88−7.76 mM 0.80−6.86 mM 1.1−38.9 mM 0.3−20 mM 17.6−1768 μM 1.1−50.0 mM 1.1−33.3 mM

6s 13 s 30 s 12 s 60 s 180 s 174 s (max) ∼ 2 min for most cartridges

LifeScan, Inc., USA

glucose lactate creatinine glucose lactate cholesterol triglycerides glucose lactate creatinine urea nitrogen glucose

Nova Biomedical, USA

5s

on the same reporter surface either requires expensive and labor intensive patterning processes to assure spatial sensitivity for the reporter elements or orthogonal transduction methods for each analyte if the recognition units are immobilized randomly. Compared to traditional designs, this problem can be more readily resolved at the nanoscale by employing nanomaterialbased reporters functionalized with different biorecognition units.77−79 Thus, instead of large sensing surface areas with specificity to only one component, such as colorimetric substrates or conventional electrodes, biosensors could be integrated with tens if not hundreds of nanoscale components that each targets an individual analyte. For example, it has been demonstrated that different types of field effect transistor (FET)-based biosensors could be combined in a way similar to integrated circuits to achieve real-time, multiplex sensing capabilities.80,81 Similarly, traditional electrodes in electrochemical biosensors can be replaced with multiple independent nanoelectrodes based on one-dimensional (1D) nanowires with high densities and multiple functionalities.82 An alternative solution includes the use of label-free nanoscale biosensors with fingerprinting capabilities. Recent advancements in plasmonic nanostructure synthesis bring new possibilities for a new kind of biosensor that can uniquely identify target analytes by directly probing their characteristic molecular properties and corresponding responses to the changing environment. Most popular among them are vibrational spectroscopic methods such as surface-enhanced Raman scattering (SERS) and surface-enhanced infrared absorption (SEIRA) spectroscopy.83,84 Additionally, advancements in optically active nanomaterials such as quantum dots with tunable adsorption features potentially remove barriers for the miniaturization of bulky optical instruments such as spectrometers.85 The second major challenge facing traditional biosensor designs is the irreversibility of antibody−antigen binding.86 This limitation does not affect solution-based chemical assays, as the reagents are washed away after biodetection events. However, for biosensors with immobilized antibodies as biorecognition units, there is no easy way to regenerate transduction surfaces. For nondisposable, large-area sensing surfaces fabricated from expensive materials such as gold or platinum, regeneration of the surface requires chemical cleaning and reimmobilization of new biorecognition units.87 Obviously, such a procedure is not feasible for rapid and continuous detection and defeats the point of having a “reagentless” sensor design. One way to solve this problem is through the creation

of scale. Additionally, the overall system accuracy of these traditional biosensors decreases as they are miniaturized. For example, when testing glucose concentrations lower than 4.2 mM, it is not uncommon for hand-held glucose sensors to show ±5% to ±10% deviation compared to results obtained from professional benchtop analyzers.69 Thus, miniature handheld devices are usually only applied to clinical analytes that have relatively high tolerance for measurement error, need fast measurements, and frequent monitoring at low cost. The current end-user needs for rapid point-of-care detection of other types of biomolecular analytes, on the other hand, remain largely unmet. These targets include, but are not limited to, various cancer biomarkers such as α-fetoprotein, prostatespecific antigen, and carcinoembryonic antigen, hormones such as testosterone, cortisol, and vasopressin, neurotransmitters such as adrenaline, serotonin and dopamine, biotoxins such as ricin, botulinum toxin, and anthrax toxin, as well as a variety of pathogens.70−75 Encouragingly, in recent years, the field of biosensors has been rejuvenated by rapid advancements in nanotechnologybased fabrication approaches and the synthesis of a variety of novel functional multiphase nanomaterials that can be readily biofunctionalzed.76 As we shall discuss, progress in the fields of nanomaterial synthesis and fabrication as well as in their assembly, nanoscale surface chemistry, and synthetic molecular recognition units has opened up new ways to drastically improve the performance of nanoscale-based biosensors and achieve commercial success for a wide range of practical applications. We briefly summarize the technological barriers current biosensing technologies are facing and discuss how nanotechnology-based approaches can provide promising ways to generally address them before delving into the principles and specifics of the nanoscale design of biosensing elements in the following sections. 2.1. Specificity and Multiplexing Capability

One of the largest difficulties with traditional biosensor designs is that due to the specificity of biorecognition units, they are typically designed to detect a single chosen analyte with specialized instrumentation. This limitation severely decreases their broad market value (except for a few analytes discussed above such as glucose and lactate that have high sensing demand) and significantly drives up the deployment cost. In comparison, traditional immunoassays can use the same set of equipment and protocols to detect a wide range of analytes, and even multiple analytes simultaneously, by swapping out different reagents such as labeled antibodies.57 In contrast, incorporating multiple different types of biorecognition units 12946

DOI: 10.1021/acs.chemrev.7b00088 Chem. Rev. 2017, 117, 12942−13038

Chemical Reviews

Review

when sensors are deployed against real-life samples and environments with much more complex chemistries and often lead to technology transfer failure.94 The adsorbed surface layer caused by excessive biofouling not only blocks access to biorecognition units, preventing quantification or even detection of analytes, but also overwhelms the transduction process in biosensors by generating “false positive” signals. However, despite decades of research on “protein resistant surfaces,” biofouling is still a major limiting factor in the reliable performance of biosensors of any type. Traditional biostealth technologies such as those utilizing polyethylene glycol (PEG) coatings can perform relatively well (