Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2022 Aug 3;14(37):42374–42387. doi: 10.1021/acsami.2c07701

Nanostructured Cyclodextrin-Mediated Surface for Capacitive Determination of Cortisol in Multiple Biofluids

Zahra Panahi †, Tianyu Ren ‡, Jeffrey Mark Halpern †,*
PMCID: PMC9504479  PMID: 35918826

Abstract

graphic file with name am2c07701_0013.jpg

The aim of this work is to develop a reusable polypropylene glycol (PPG):β-cyclodextrin (βCD) biosensor for cortisol detection. To achieve the most stable support for βCD, we developed two PPG surfaces. The first surface is based on a gold surface modified with SAM of 3-mercaptopropionic acid (3MPA), and the second surface is based on a glassy carbon surface grafted with 4-carboxyphenyl diazonium salt. We characterized both surfaces by EIS, XPS, and ATR-FTIR and evaluated the stability and reusability of each surface. We found the GC-carboxyphenyl-PPG:βCD is stable for at least 1 month. We have also demonstrated the reusability of the surface up to 10 times. In detecting cortisol, we used a nonfaradaic electrochemical impedance capacitive model to interpret the surface confirmation changes. We achieved sensitive detection of cortisol in PBS buffer, urine, and saliva with limit of detection of 2.13, 1.29, and 1.33 nM, respectively.

Keywords: electrochemical, electrochemical impedance spectroscopy, cortisol, biosensor, reusable surface, biotechnology, surface stability

1. Introduction

Sepsis is a combination of infection and the systemic response to that infection. Increased energy demands in sepsis stimulates gluconeogenesis processes and induce secretion of stress hormones. Therefore, elevated levels of stress hormones such as cortisol have been demonstrated in sepsis.1 Unlike some other sepsis biomarkers that have a short half-life, cortisol is a stable sepsis biomarker.2 A study that observed and analyzed inflammatory parameters of 62 intensive care unit (ICU) patients demonstrated a significant positive correlation between blood cortisol concentrations and daily sequential organ failure assessment (SOFA) scores.3 However, cortisol levels change rapidly and measurement of blood cortisol concentrations requires the blood test which may increase stress hormone levels.4 Therefore, measurement of cortisol in noninvasive relevant biofluids such as urine and saliva can yield better data.5 Septic patients typically will see an increase of urinary cortisol of 3–8 times, and salivary cortisol levels in severe sepsis patients is 2.6 times higher than healthy subjects.1,6

Various electrochemical,7−13 colorimetric,4,14 and surface plasmon resonance15 methods have been used to develop cortisol biosensors. Despite the progress made in recent years, most of the research in this area has focused on employing different methods to immobilize antibodies or aptamers on the surface. However, a recent study has shown that compared to the reference method of high performance liquid chromatography (HPLC), biosensors based on antibody or aptamer recognition mechanisms can have a wide range of interassay and intraassay variation leading to an overestimate or underestimate of the true cortisol concentrations.16 Therefore, development of a chemically modified reusable biosensor would provide a potential tool for clinicians to continuously measure the cortisol levels and monitor the course and severity of sepsis in hospitalized patients. To our knowledge, no reusable biosensor for continuous cortisol measurements has ever been developed and all the current cortisol biosensors are single use.

Cyclodextrins, with a hydrophobic cavity and a hydrophilic outer surface, have many applications as a solubilizer, stabilizer, and carrier in pharmaceutical industry.17,18 Our group has previously developed a reusable α-cyclodextrin (αCD) based biosensor for detection of resveratrol in urine,19 but αCD does not bind to many biomarkers, β-cyclodextrin (βCD) and cortisol can make host/guest inclusion complexes, and βCD has been used as a carrier for cortisol in various drug delivery systems.20,21 Additionally, the −OH groups of βCD allow hydrogen bonding with polymer materials such as polypropylene glycol (PPG) derivatives. The molecular weight of the PPG main chain and the size and shape of the PPG end group affect the efficiency of the complex formation between βCD and PPG. For example, βCD can form channel type crystallin complexes with PPG (MW = 2000) with NH2 end groups because the NH2 groups are small enough to pass through βCD molecules.22,23

In the previous study, a gold electrode modified with a thiol containing polyethylene glycol (thiol-PEG) was used as a support for αCD.19 We showed that the PEG:αCD surface could be regenerated three times; however, the surface lost its stability after four or five uses. We hypothesize this limitation was imposed by low stability of the covalent bond between gold and thiol-PEG. The stability of the covalent bonds can be improved by different methods such as synthesis of multithiol or N-hetrocyclic carbene (NHC) functionalized polymer supports.24,25 Alternatively, surface-immobilization can be performed by reduction of diazonium salts.26−28

To investigate our hypothesis and address the problem of low stability of cyclodextrin support, in this paper, we compared the stability of gold–thiol versus carbon–carbon covalent bonds to develop the most stable PPG:βCD biosensor. The gold-PPG surface (Scheme 1A) was fabricated based on thiol self-assembled monolayers (SAMs) of 3MPA. Adsorption of 3MPA thiols on a gold surface produces ordered monolayers, and the carboxylic acid group attached to the chain allows further modification with PPG. The second PPG surface (Scheme 1B) was prepared based on carbon–carbon covalent bond between glassy carbon and aryl diazonium salt. Among all diazonium salts with various functional groups, 4-carboxyphenyl diazonium salt with a carboxylic acid group was chosen to make the two surfaces comparable. Therefore, carboxylic acid groups of both 3MPA and carboxyphenyl surfaces could be activated and modified with NH2-PPG-NH2 via carbodiimide/succinimide (EDC/NHS) chemistry (Scheme 1).

Scheme 1. Surface Modification Steps of (A) Gold-3MPA-PPG:βCD and (B) GC-carboxyphenyl-PPG:βCD.

Scheme 1

Open in a new tab

The aims of this paper are, first, to develop and characterize two PPG surfaces to find the most stable and reusable βCD biosensor and, second, to investigate the performance of the biosensor for nonfaradaic electrochemical impedance spectroscopy (EIS) measurement of cortisol in biofluids (i.e., urine and saliva). While extensive research has been conducted to analyze faradaic EIS data,29 some problems still remain with the analysis of nonfaradaic EIS. In a typical nonfaradaic EIS measurement, the parameters associated with electron transfer such as charge transfer resistance and Warburg impedance are eliminated, and this makes finding the equivalent electrical circuit (EEC) that fits the data challenging.30,31 In this work, instead of EEC based analysis, we employed a different method that involves measurement of imaginary capacitance at various frequencies.

2. Experimental Section

2.1. Chemicals and Reagents

Sodium nitrite (Avocado Research Chemicals >95%), 4-aminobenzoic acid (Alfa Aesar 99%), hydrochloric acid (HCl) (Sigma-Aldrich 37%), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC) (Thermo Scientific 100%), N-hydroxysuccinimide (NHS) (Thermo Scientific 100%), poly(propylene glycol)bis(2-aminopropyl ether) (Sigma-Aldrich 100%), ethanolamine (Alfa Aesar 100%), β-cyclodextrin (Acros Organics 98%), 3-mercaptopropionic acid (Acros Organics 99%), ethanol (ACS Reagent), potassium ferrocyanide (Acros Organics, ACS Reagent), potassium ferricyanide (Acros Organics, 98%), hydrocortisone (98% Acros Organics), trans-resveratrol (Sigma-Aldrich 99%), acetaminophen (Sigma-Aldrich), uric acid (99% Alfa Aesar), and artificial saliva (Pickering Laboratories In (pH 6.8)) were purchased.

PBS was prepared using a standard protocol from ThermoFisher Scientific (pH 7.4). Raw, unprocessed pooled human urine from 20 donors was provided by Lee BioSolutions (pH 6.11 and conductivity 13.06 mS/cm).

BASi gold and glassy carbon (GC) disk electrodes (area = 0.0201 cm2), gold-coated silicon wafers, and screen-printed carbon electrodes were purchased from BASI, TED PELLA, and Metrohm DropSens, respectively.

Ultra high purity (UHP) water is purified with a reverse osmosis and deionized with a Millipore Direct Q3 system and measured at 18.2 MΩ.

2.2. Gold-3MPA Modification

A BASi gold disk electrode was polished in a figure-eight pattern by 1 μm diamond for 1 min and then it was rinsed with methanol and UHP water.19 The electrode was then polished with 0.55 μm alumina for 1 min, and it was sonicated for 1 min in UHP water to remove residual alumina particles. The electrode was dried with nitrogen, and oxygen plasma was used to remove any remaining organic contaminations. The gold surface was then modified by 10 mM 3-mercaptopropionic acid (3MPA) in ethanol.32 The gold surface was incubated in the 3MPA solution for 24 h to allow thiol groups to covalently bond with gold surface. The electrode was then rinsed thoroughly with UHP water to remove physisorbed molecules and dried with N2 prior to PPG modification.

2.3. In Situ Generation of 4-Carboxyphenyl Diazonium Salt and Electrografting of 4-Carboxyphenyl Diazonium Salt on Glassy Carbon

A BASi glassy carbon disk electrode was polished with 0.55 μm alumina for 1 min, and the electrode was then sonicated in UHP water for 1 min and rinsed thoroughly with UHP water. In situ generation of 4-carboxyphenyl diazonium salt was performed by the reaction between 4-aminobenzoic acid (C7H7NO2), sodium nitrite (NaNO2), and hydrochloric acid (HCl).28 Since this reaction is exothermic, the reaction vessel was placed in an ice bath to control the temperature. First, 10 mL of 2 mM 4-aminobenzoic acid was added to 10 mL of 0.5 M hydrochloric acid. Then 10 mL of 2 mM sodium nitrite was slowly added and the solution was stirred for 10 min. 15 mL of 4-carboxyphenyl diazonium salt solution was transferred into a shot glass and the polished glassy carbon surface was immersed in the solution. The electrochemical grafting of glassy carbon surface was performed with Gamry Reference 600+ potentiostat. Three cyclic voltammetry (CV) scans were run to graft 4-carboxyphenyl diazonium salt on the surface. CV scans were conducted at 100 mV/s by potential cycling between 0.4 and −0.6 V vs Ag/AgCl. The electrode was then rinsed with UHP water and dried with N2 prior to the next step. The remaining 4-carboxylphenyl diazonium salt solution was quenched with hypophosphorous acid (H3PO2) immediately after modification of the electrodes.33

2.4. PPG Attachment on Gold-3MPA and GC-carboxyphenyl Surfaces

The carboxylic groups are present on both gold-3MPA and GC-carboxyphenyl surfaces. EDC/NHS chemistry was used to activate carboxylic groups. 10 mL of 100 mM EDC/20 mM NHS in MES buffer (pH = 5) was prepared, and the electrodes were incubated in the solution for 90 min. Then the electrodes were rinsed with UHP water and modified with poly(propylene glycol)bis(2-aminopropyl ether). 5 mg of PPG was added to 20 mL of PBS and stirred by sonication for 10 min. Gold-3MPA-PPG and GC-carboxyphenyl-PPG surfaces were prepared by immersing the electrodes in the PPG solution for 30 min. The electrodes were rinsed with UHP water followed by incubating the surfaces in 1 M ethanolamine solution in PBS for 10 min to deactivate excess reactive groups.34 Finally, the electrodes were rinsed with UHP water and dried with N2 and stored in UHP water.

2.5. Surface Characterization (EIS, ATR-ATR-FTIR, and XPS)

All electrochemical measurements were performed with Gamry Instruments Reference 600+ potentiostat in aC3 Cell Stand. A three electrode system comprising an Ag/AgCl/3 M NaCl reference electrode, a platinum wire auxiliary electrode, and a gold or glassy carbon disk electrode working electrode was used for electrochemical experiments. These measurements were performed within the frequency range of 0.1 Hz to 100 kHz, acquiring 10 points per decade with a DC potential of 0 V versus open-circuit potential and an AC voltage of 10 mV. The faradaic EIS measurements for surface characterization and monitoring of the surface stability were carried out in 20 mL of 20 mM ferri/ferrocyanide in UHP.19 The faradaic data were fitted with Randles circuit to find charge transfer resistance Rct for different surfaces. Nonfaradaic EIS was conducted in the indicated buffer without a supporting electrochemical mediator. Nonfaradaic EIS was interpreted through the capacitive components of the impedance.

ATR-FTIR measurements were performed with Thermo Nicolet (iS10 ATR-FTIR), and XPS data collection was conducted using Kratos AXIS Supra XPS at the university instrumentation center of University of New Hampshire. To prepare samples for these tests, gold-coated silicon wafers and screen-printed carbon electrodes were used to prepare gold-3MPA-PPG and GC-carboxyphenyl-PPG, respectively. All samples were rinsed with UHP water prior to ATR-FTIR and XPS characterization.

2.6. Long-Term Stability of Gold-3MPA-PPG and GC-carboxyphenyl-PPG

Gold-3MPA-PPG (N = 3) and GC-carboxyphenyl-PPG (N = 3) modified surfaces were stored in UHP water at room temperature, and an EIS measurement in 20 mM ferri/ferrocyanide aqueous solution was performed every 7 days for up to a month.

2.7. NMR Studies on Complex Formation of βCD with PPG and Cortisol

Since both PPG and cortisol are barely soluble in water, we added a small amount of methanol to dissolve these compounds for NMR studies. 10 mg of PPG was dissolved in 1 mL of methanol. The solution was then mixed with 5 mL of saturated βCD solution (containing 90 mg of βCD in 5 mL of UHP water). The βCD-PPG solution was sonicated for 15 min and allowed to stay overnight at room temperature. The white powder precipitates were collected by centrifugation and vacuum-dried. The powder was then characterized by 1H NMR (DMSO-d6, 500 MHz).35 Similarly, 10 mg of cortisol was dissolved in 1 mL of methanol, and 7 mL of saturated βCD solution (containing 126 mg of βCD in 7 mL of UHP water) was added. The solution was sonicated for 15 min and stayed overnight. The white powder precipitates were collected by centrifugation, dried by vacuum, and characterized by 1H NMR. Finally, PG:βCD:cortisol complexes were characterized by 1H NMR. To prepare the samples, 10 mg of PPG and 10 mg of cortisol were dissolved in 1 mL of methanol. This solution was mixed with 7 mL of saturated βCD solution (containing 126 mg of βCD in 7 mL of UHP water). The mixture was sonicated for 15 min and allowed to stay overnight at room temperature. The product was then collected by centrifugation and vacuum-dried.

2.8. βCD Loading and Investigating the Reusability of Surfaces

Gold-3MPA-PPG and GC-carboxyphenyl-PPG surfaces were immersed in 5 mM βCD aqueous solution for 1 h to prepare gold-3MPA-PPG:βCD and GC-carboxyphenyl-PPG:βCD surfaces, respectively. The surfaces were washed thoroughly with UHP, and an EIS run was performed in PBS to measure imaginary capacitance. Then the surfaces were incubated in ethanol for 1 h to remove βCDs from the surface. The surfaces were rinsed with UHP, and imaginary capacitance was measured. We hypothesized that the surfaces could then be reloaded with another 1 h soak in 5 mM βCD solution. To investigate our hypothesis and evaluate the reusability of surfaces, independent gold-3MPA-PPG (N = 3) and GC-carboxyphenyl-PPG (N = 3) were modified and the process of βCD loading/ethanol soak was conducted 10 consecutive times for each surface (n = 10).

Throughout this manuscript we use N to indicate the number of different surfaces or electrode preps and n to indicate multiple runs from the same surface. Both N (reproducibility of electrodes, interassay reproducibility) and n (reusability of the same electrode, intraassay reproducibility) are important for this study.

2.9. Assessment of the Sensing Stability in PBS during EIS Measurements

To determine measurement drift, after βCD loading (1 h), the electrodes were immediately used for nonfaradaic EIS experiments in PBS. Imaginary capacitance was measured over a frequency spectrum of 0.1–105 Hz. The normalized imaginary capacitance change (Inline graphic) at all frequencies was calculated for gold-3MPA-PPG:βCD (N = 4) and GC-carboxyphenyl-PPG:βCD (N = 5). C″ is the imaginary capacitance at t = 0 ,and C″ is the imaginary capacitance at t = 30 min.

2.10. Cortisol Detection and Control Experiments in PBS, Urine, and Artificial Saliva

To determine the baseline (blank solution with no cortisol), the GC-carboxyphenyl-PPG:βCD surface was first incubated in blank PBS and an EIS was run after 5 min. Then, the surface was immersed in the cortisol solution, with the desired concentration, for 5 min and then EIS was run. The normalized imaginary capacitance change (Inline graphic) was used to plot calibration curves in PBS, urine, and saliva. C″0 is the imaginary capacitance of baseline (blank solution), and C″ is the imaginary capacitance of concentrated cortisol solution.

The concentrated cortisol stock solutions in PBS (pH 7.4), urine, and saliva were prepared. For PBS experiments, serial dilutions of cortisol solutions were done to get 2.5–160 nM concentrations. For urine and saliva experiments, serial dilutions of cortisol solutions were done to get 2.6–168.6 nM concentrations. For control experiments in PBS, urine, and saliva, an equivalent volume of blank solutions was added to the test solutions because the dilution of the solution can decouple the CD from the surface.

2.11. Evaluation of GC-carboxyphenyl-PPG:βCD Selectivity in the Presence of Potential Interfering Analytes

The GC-carboxyphenyl-PPG:βCD surface was first immersed in PBS, and after 5 min, an EIS was run to determine the baseline value (C″0). Then different volumes of concentrated solutions were added to get 160 nM trans-resveratrol, uric acid, and acetaminophen solutions (C″). Then, the normalized imaginary capacitance change (Inline graphic) was calculated for different analytes.

3. Results and Discussion

3.1. Gold-3MPA-PPG and GC-carboxyphenyl-PPG Characterization

Selection of a robust and stable support is an important step in developing reusable biosensors. In an attempt to develop a reusable βCD platform, in this work we evaluated two different PPG surfaces, as a βCD support, to evaluate the long-term stability and reuse of each surface. The first surface (gold-3MPA-PPG:βCD) was constructed through covalent modification of 3MPA on gold via a thiol bond. The second surface (GC-carboxyphenyl-PPG:βCD) was glassy carbon modified by reduction of in situ electrochemically generated 4-carboxyphenyl diazonium salt. Grafting onto the glassy carbon electrode was performed by cyclic voltammetry (CV), and the first scan showed a cathodic peak at −0.25 V vs Ag/AgCl reference electrode as shown in Figure 1. This cathodic peak resulted from loss of a N2 molecule, formation of an aryl radical, followed by covalent binding of the aryl radical to the glassy carbon surface.28 Disappearance of the cathodic peak in the second and third CV scans indicates the glassy carbon surface is fully modified with carboxyphenyl groups.

Figure 1.

Figure 1

Open in a new tab

Cyclic voltammograms of electrochemical grafting of carboxyphenyl groups on GCE in 4-carboxyphenyl diazonium salt solution from +0.4 to −0.6 vs Ag/AgCl at scan rate of 100 mV s–1. The characteristic cathodic peak in the first CV scan suggests the formation of a carboxyphenyl layer. Due to the blocking effect of the carboxyphenyl layer, the cathodic peak disappears in the second and third CV scans.

We studied surface modification of both surfaces using EIS, ATR-FTIR, and XPS. The results of the faradic EIS in 20 mM ferri/ferrocyanide are demonstrated in Figure 2. The insets in Figure 2a and Figure 2b show the Nyquist plots of an unmodified gold (orange circle) and an unmodified glassy carbon (black circle), respectively. The diameter of the semicircle in Nyquist plot is directly proportional to the charge transfer resistance. An increase in charge transfer resistance is observed by attachment of 3MPA (green triangle) and carboxyphenyl (blue triangle) followed by modification of both surfaces with PPG (red square) due to the additional passivation of electrode. We infer from these graphs that we successfully modified both PPG surfaces.

Figure 2.

Figure 2

Open in a new tab

Nyquist plots of faradaic EIS in 20 mM ferri/ferrocyanide in UHP: (A) a gold surface before and after modification with 3MPA and PPG and (B) a glassy carbon surface before and after modification with carboxyphenyl and PPG. The increase in diameter of the semicircle in Nyquist plots indicates an increase in the charge transfer resistance.

In order to confirm successful PPG modification on electrode surfaces, we analyzed the functional groups of gold-3MPA-PPG and carbon-carboxyphenyl-PPG surfaces by ATR-FTIR (Figure 3A and Figure 3B). According to Figure 3A, two peaks at 1239 and 1400 cm–1 are noticed for 3MPA spectrum (green). These peaks correspond to C–O and belong to aliphatic chain of 3MPA. The peak at 1695 cm–1 is assigned to the carboxylic acid end group of 3MPA. However, this peak disappears after PPG modification (red), suggesting the binding interaction of PPG to the carboxylate on top of the 3MPA surface. The peaks at 1100, 1376, and 1450 are due to C–O stretch, and the peaks at 2865 and 2972 cm–1 arise from C–H stretching groups of PPG.36 In Figure 3B, the presence of carboxyphenyl groups of diazonium salt can be confirmed by peaks observed at 1520 and 1660 cm–1 (blue). These peaks may be attributed to the aromatic and/or carboxylic acid groups, and they vanish after the attachment of PPG (red).37 The C–O stretching peaks of PPG appear in the 1000–1500 cm–1 range, and the 2865 and 2972 cm–1 peaks indicate the presence of C–H groups of PPG.36

Figure 3.

Figure 3

Open in a new tab

ATR-FTIR and XPS characterization of surfaces: (A) ATR-FTIR of gold-3MPA-PPG, (B) ATR-FTIR of carbon-carboxyphenyl-PPG, (C) XPS of gold-3MPA-PPG, and (D) XPS of carbon-carboxyphenyl-PPG. The ATR-FTIR peaks at 1239 and 1400 cm–1 in (A) are assigned to C–O, and the peak at 1695 cm–1 belongs to C=O binding of 3MPA. The peaks at 1520 and 1660 cm–1 in (B) can be attributed to the aromatic and/or carboxylic acid groups of carboxyphenyl surface. In both 3MPA and carboxyphenyl surfaces in (A) and (B), after PPG modification, C–H stretching groups (at 2865 and 2972 cm–1) of PPG dominate the spectra and mask 3MPA and carboxyphenyl signals. The comparison of XPS data after different modification steps in (C) and (D) shows successful modification of gold-3MPA-PPG and carbon-carboxyphenyl-PPG surfaces.

XPS analysis was performed to verify the EIS and ATR-FTIR results (Figure 3C and Figure 3D). The orange graph in Figure 3C shows an unmodified gold surface with dominant gold (Au 4f) peaks. After modification of gold surface with 3MPA (green), the intensity of the gold peak decreases and some carbon (C 1s) and oxygen (O 1s) peaks appear. The presence of a sulfur peak in the 3MPA spectrum is due to the thiol-gold bonding. Appearance of nitrogen (N 1s) peak after PPG modification can be assigned to unreacted NH2 groups of PPG. Table S1 shows the decrease in Au/C ratio from 122.1 to 18.8 by 3MPA modification and increase in C/O from 1.75 to 5.37 by PPG modification. Wide XPS spectrum of carbon-carboxyphenyl-PPG surface is shown in Figure 3D. The presence of carboxyphenyl groups and PPG molecules is supported by the increase in C 1s and O 1s signals after modification of carbon surface with diazonium salt (blue) and PPG (red), respectively. Also, the N 1s peak appears upon modification of diazonium surface with PPG. Finally, Table S2 indicates that by modification of carboxyphenyl surface with PPG, the C/O ratio increases from 0.63 to 4.14.

3.2. Long-Term Stability of Gold-3MPA-PPG and GC-carboxyphenyl-PPG Surfaces

EIS data are usually interpreted with the use of equivalent electrical circuits. The most common electrical circuit to fit faradaic EIS data is the Randles circuit (Figure 4A). The Randles circuit consists of a solution resistance Rs in series with the double layer capacitance Cdl, charge transfer resistance Rct, and Warburg impedance. Double layer capacitance deviates from ideal behavior due to surface roughness, and constant phase element (CPE) is often used in the equivalent circuit model. We used the Randles circuit to fit the faradaic EIS data shown in Figure 2.38 The charge transfer resistance values for gold-3MPA-PPG and GC-carboxyphenyl-PPG surfaces are represented in Figure 4B and Figure 4C, respectively (average and standard deviation reported, N = 3). To assess the long-term stability of modified PPG surfaces, we kept both surfaces in room temperature UHP water and measured the charge transfer resistance weekly. While the charge transfer resistance of the gold-3MPA-PPG surface decreased dramatically after 1 week (p < 0.05), the charge transfer resistance of the GC-carboxyphenyl-PPG surface remained constant (no statistically relevant changes). These data indicate that stability GC-carboxyphenyl-PPG surface is higher than the gold-3MPA-PPG, and the replacement of thiol bonds by C–C contributes to the higher stability of biosensor.

Figure 4.

Figure 4

Open in a new tab

Surface modification and stability test of gold-3MPA-PPG and GC-carboxyphenyl-PPG surfaces for a period of 1 month. Electrodes were stored in UHP water between stability tests; stability testing was conducted in 20 mM ferri/ferrocyanide in UHP. (A) Randles circuit that was used to fit the faradaic EIS data. (B) Average charge transfer resistance (Rct) for three gold-3MPA-PPG surfaces. (C) Average charge transfer resistance for three GC-carboxyphenyl-PPG surfaces. Error bars indicate the standard deviation of three independent surfaces (N = 3). The data indicate that the GC-carboxyphenyl-PPG surface is more stable than gold-3MPA-PPG after just 1 week (*p < 0.05).

3.3. Capacitive Modeling and Reusability of Surface Construction

βCD forms inclusion complexes with PPG if the molecular weight is 2000, and a single βCD molecule is estimated for every three propylene glycol (PG) units.35 We confirmed this estimate with 1H NMR, and the spectra of PPG, βCD, and PG:βCD (3:1) are shown in Figure S1. 1H NMR was also used to confirm βCD:cortisol binding (Figure S2). Our data demonstrate that every two βCD molecules bind with one cortisol molecule (2:1). Additionally, Figure S3 shows the PG:βCD:cortisol (6.3:4.5:1) inclusion complexes. Table S3 summarizes the results from Figures S1–S3 and shows the stoichiometric ratio of βCD:PG and βCD:cortisol in the absence and presence of cortisol and PPG, respectively. When there is only cortisol and βCD in the solution and there is no PPG present, CD:cortisol = 1:0.5 and when there is no cortisol in the PPG and βCD solution, the CD:PG = 1:3. However, when we have CD, cortisol, and PPG in the solution, CD reacts with both cortisol and PPG. Therefore, when cortisol is added to the βCD and PPG solution, less PG units interact with βCD and each βCD molecule interacts with 1.4 PG units instead of 3 PG units. This makes sense because the remaining βCD molecules interact with cortisol. Similarly, when PPG is added to the βCD and cortisol solution, less cortisol units interact with βCD and each βCD molecule interacts with 0.22 cortisol molecule instead of 0.5 cortisol molecule. Again, this makes sense because βCD interacts with both PPG and cortisol and the remaining βCD molecules interact with PG units.

Charge transfer resistance is often used to interpret the faradaic EIS in biosensors. However, the use of redox couples limits the practical applications of biosensors in complex biological fluids.39 Redox couples, such as ferri/ferrocyanide, can interact with surface or analytes and deteriorate surface sensitivity and selectivity. Therefore, a nonfaradaic EIS approach, in the absence of ferri/ferrocyanide, should be used with cyclodextrin-based biosensors.

The capacitance changes in nonfaradaic EIS tend to be small which requires a perfectly fit circuit to monitor and predict these changes. An alternative to fitting the data to an equivalent circuit is to use total resistance or capacitance data at a fixed angular frequency ω. The total capacitance C is represented by the real part C′ and the imaginary part C″, and it is expressed as follows:40

3.3. 1
3.3. 2
3.3. 3

where Z′ and Z″ are real and imaginary parts of the impedance, respectively. In this work, the imaginary part of the capacitance C″ (eq 3) has been used to interpret the EIS output of the sensor since the imaginary part displayed a higher sensitivity than the total capacitance (eq 1). Figure 5 demonstrates the short-term (sensing) stability of gold-3MPA-PPG:βCD and GC-carboxyphenyl-PPG:βCD surfaces in blank PBS over 30 min. A comparison between Figure 5A and Figure 5B shows that the imaginary capacitance change for gold-3MPA-PPG:βCD in blank PBS is smaller than GC-carboxyphenyl-PPG:βCD. The sensing stability of GC-carboxyphenyl-PPG:βCD surface was proved to be higher than gold-3MPA-PPG:βCD; therefore the diazonium mediated surfaces are more appropriate for sensing applications over thiol surfaces.

Figure 5.

Figure 5

Open in a new tab

Comparison between the stability of surfaces in PBS from t = 0 to t = 30 min: (A) gold-3MPA-PPG:βCD (N = 4) and (B) GC-carboxyphenyl-PPG:βCD (N = 5). These data show that GC-carboxyphenyl-PPG:βCD has less imaginary capacitance change and higher sensing stability than gold-3MPA-PPG:βCD surface.

The reusability experiments were performed by developing three surface replicates of gold-3MPA-PPG (Figure 6A, N = 3) and GC-carboxyphenyl-PPG (Figure 6B, N = 3). The surfaces were loaded with βCD and then soaked in ethanol to remove βCD molecules from the surface; this process was repeated 10 times measuring the imaginary capacitance after each stage. The reported values (Figure 6) are the average and standard deviation over the three independent surfaces. After each ethanol soak (i.e., βCD removal) the imaginary capacitance increased, and the imaginary capacitance decreased after βCD reloading. This result suggests that the GC-carboxyphenyl-PPG:βCD surface can be regenerated and reused up to 10 times. In addition, smaller error bars of GC-carboxyphenyl-PPG:βCD suggest higher measurement and surface-to-surface construct precision.

Figure 6.

Figure 6

Open in a new tab

Comparison between the reusability of surfaces (N = 3): (A) reusability of gold-3MPA-PPG:βCD, 10 uses from the same electrode, and (B) reusability of GC-carboxyphenyl-PPG:βCD, 10 uses from the same electrode. The comparison between (A) and (B) shows that GC-carboxyphenyl-PPG:βCD is more reproducible (smaller drift and run-to-run variation) and can be reused and regenerated up to 10 times.

3.4. Detection of Cortisol

Figure 7A schematic illustrates the biosensor platform for cortisol detection. When cortisol is introduced into this system, the βCD will depart from the PPG support because of the competitive inclusion interaction between cortisol and βCD, leading to the decrease in impedance and increase in capacitance. To confirm our theory that the increase in capacitance is due to the release of βCD from the surface, XPS was used. Elemental composition of βCD surface before and after cortisol detection was estimated by comparing C 1s (Figures 7B and S4) and O 1s (Figures 7C and S5) spectra. XPS was performed on the following four samples:

  • 1.

    Carbon-carboxyphenyl-PPG:βCD surface (βCD loaded 1)

  • 2.

    Carbon-carboxyphenyl-PPG:βCD surface after 1 h exposure to 160 nM cortisol solution (cortisol 1)

  • 3.

    Regenerated carbon-carboxyphenyl-PPG:βCD surface (βCD loaded 2) by soaking the electrode in 5 mM βCD solution for 1 h

  • 4.

    Regenerated carbon-carboxyphenyl-PPG:βCD surface after 1 h exposure to 160 nM cortisol solution (cortisol 2)

Figure 7.

Figure 7

Open in a new tab

Demonstration of GC-carboxyphenyl-PPG:βCD surface for reusable sensing of cortisol: (A) schematic that demonstrates (left) release of βCD from PPG to interact with cortisol and (right) surface regeneration by βCD reloading, (B) mass concentration (%) of C 1s peaks, and (C) O 1s peaks interpreted from XPS performed on βCD surface before and after cortisol detection, (D) mass concentration (%) of C 1s peaks, and (E) O 1s peaks interpreted from XPS performed on βCD surface before and after ethanol soak. These data indicate that βCD can be removed from the GC-carboxyphenyl-PPG surface by immersion in the cortisol solution (B, C) or ethanol soak (D, E).

The C 1s spectrum (Figure S4) can be deconvoluted into three peaks at 284.8, 286.4, and 288.5 eV which can be attributed to C–C or C–H, C–OH or C–O–C, and C=O, respectively. The bar graph in Figure 7B represents mass concentrations (%) of peaks. All three C 1s components were observed for all samples. Compared to βCD, PPG contains higher amount of C–C and C–H bonds; therefore, the increase in mass concentration of these groups from 31.72% in βCD loaded 1 surface to 62.14% in cortisol 1 surface indicates the removal of some βCD molecules followed by the incubation of surface in cortisol solution. After βCD reloading (βCD loaded 2), the mass concentration of C–C and C–H groups decreases to 38.6%, suggesting the biosensor can be successfully regenerated by βCD reloading. Moreover, the mass concentration of dominant groups of βCD, i.e., C–OH and C–O–C, decrease after the surface is exposed to cortisol solution.

For O 1s spectrum (Figure S5), we can assign the peak with 531.2 eV binding energy to C=O and the components at 532.5 and 533.5 eV can be assigned to C–O–C and OH groups. Figure 7C demonstrates the mass concentration (%) of different components. By comparing the mass concentration of C–O–C and OH components of different samples, it was found that the changes in the amount of OH groups are more prominent than C–O–C groups since C–O–C groups are present in both βCD and PPG molecules. The decrease in OH groups after the exposure of surface to cortisol (from 34.8% to 12.3%) confirms release of βCD, and the increase in OH groups after reloading of βCD (from 12.3% to 28.9%) confirms the reusability of surface.

Another set of XPS experiments on GC-carboxyphenyl-PPG:βCD surfaces were performed to show the effect of ethanol soak on surfaces. Figures 7D and S6 represent C 1s and Figures 7E and S7 represent O 1s deconvolution peaks of GC-carboxyphenyl-PPG:βCD surfaces before and after ethanol soak, after βCD reloading, and after second ethanol soak. These results are consistent with Figure 6B, suggesting βCD can be removed from the surface by ethanol soak instead of exposure to cortisol solution.

The nonfaradaic EIS response of GC-carboxyphenyl-PPG:βCD to 160 nM cortisol is shown in Figure 8A. To perform these experiments, GC-carboxyphenyl-PPG:βCD surfaces (N = 5) were tested in PBS solution with spiked cortisol solutions. Additionally, to investigate the effect of blank PBS on the biosensor response, control experiments were carried out by addition of equivalent volumes of blank PBS to the test solution (N = 5). Figure 8B shows an imaginary capacitance change in response to 122.09 μL of blank PBS solution added to the test solution. The capacitance change in blank solution can be due to the release of βCD molecules into the blank solution41 or drift in EIS measurements.42 From Figure 8A and Figure 8B, the normalized imaginary capacitance change of a frequency with the highest signal-to-noise ratio (SNR) was selected as the biosensor signal. The SNR was defined by eq 4 and calculated for test (cortisol) and control (PBS).

3.4. 4

where μ is the mean and δ is standard deviation. The ratio of the SNR of test to control was calculated at different frequencies, and 1577 Hz, which had the highest SNR between test to control, was selected as the optimal frequency to evaluate the normalized imaginary capacitance change (Figure 9A and Figure 9B).

Figure 8.

Figure 8

Open in a new tab

Comparison between the GC-carboxyphenyl-PPG:βCD response to (A) 160 nM cortisol solution (122.09 μL of concentrated cortisol solution added to PBS test solution) (N = 5) and (B) 122.09 μL of blank PBS solution added to PBS test solution (N = 5). For all frequencies, the ratio of (SNR of (A)/SNR of (B)) was calculated to find the frequency with highest value to find the best frequency to interpret the data.

Figure 9.

Figure 9

Open in a new tab

Cortisol detection in PBS using GC-carboxyphenyl-PPG:βCD. (A) Imaginary capacitance vs frequency plot for various cortisol concentrations. (B) The normalized imaginary capacitance change has the highest SNR at 1577 Hz. (C) Biosensor response to cortisol (black squares) and blank PBS control (red circles). These data show that by increasing cortisol concentration, the imaginary capacitance change at 1577 Hz increases and was only due to the interaction of βCD with cortisol.

The normalized imaginary capacitance change of GC-carboxyphenyl-PPG:βCD as a function of cortisol concentration at 1577 Hz is plotted to generate a calibration curve in Figure 9C. The results show that the imaginary capacitance change response increases with increasing cortisol concentration in the range of 2.5–160 nM, and the data can be fit with a Langmuir isotherm. This indicates that the binding affinity of βCD and cortisol must be higher than the binding affinity of βCD and PPG. Therefore, once the βCD surface is introduced to the cortisol solution, the βCD dissociates from PPG to interact with cortisol in the solution (eqs 5–7).

3.4. 5
3.4. 6
3.4. 7

where Q is biosensor signal (Inline graphic), N is the maximum capacity of βCD:cortisol interactions, and Kd is dissociation constant.43−45

The Kd of cortisol calibration curve in PBS (Figure 9C, black squares) was determined to be 0.18 nM–1, and N was calculated to be 5.56. On the basis of eqs 8 and 9,46 the limit of blank (LOB) and limit of detection (LOD) were calculated to be 0.94 and 2.13 nM, respectively.

3.4. 8
3.4. 9

Finally, the biosensor’s response to increasing volumes of blank PBS (Figure 9C, red circles) cannot be fitted by Langmuir isotherm, confirming that the blank PBS has almost no influence on imaginary capacitance response of cortisol at 1577 Hz. Therefore, increasing concentrations of cortisol was solely responsible for the increase in imaginary capacitance response.

3.5. Selectivity of GC-carboxyphenyl-PPG:βCD in the Presence of Potential Interfering Analytes

In order to assess potential interfering analytes for clinical applications with the GC-carboxyphenyl-PPG:βCD sensor, several compounds including uric acid, acetaminophen, and resveratrol were evaluated (Figure 10). Uric acid is one of the most important coexisting analytes in urine, and these data suggest that uric acid does not interfere with sensing of cortisol, and the biosensor can be tested in urine samples (p < 0.001). Additionally, contrary to interfering effect of acetaminophen with a previously developed βCD biosensor,47 GC-carboxyphenyl-PPG:βCD does not respond to acetaminophen (p < 0.005). Finally, while resveratrol can make host–guest inclusion complexes with αCD, resveratrol does not influence the cortisol detection by the developed βCD surface (p < 0.005). In summary, the normalized imaginary capacitance change in response to cortisol was significantly greater than blank PBS, uric acid, acetaminophen, and resveratrol (p < 0.005). However, pairwise comparisons between uric acid and acetaminophen, uric acid and resveratrol, uric acid and blank PBS, acetaminophen and resveratrol, acetaminophen and blank PBS, and resveratrol and blank PBS showed these were not significantly different from each other.

Figure 10.

Figure 10

Open in a new tab

Comparison of GC-carboxyphenyl-PPG:βCD biosensor response to cortisol (160 nM) vs blank PBS and potential interferents including uric acid (160 nM), acetaminophen (160 nM), and resveratrol (160 nM) in PBS. The data confirm that uric acid, acetaminophen, and resveratrol do not interfere with cortisol analysis (p < 0.005).

3.6. Performance of GC-carboxyphenyl-PPG:βCD for Detection of Cortisol in Urine and Saliva

We evaluated the capability of our modified GC-carboxyphenyl-PPG:βCD surface to identify cortisol spiked pooled urine and artificial saliva samples. Similar to Figure 9C, the Langmuir isotherm was used to describe desorption of βCD molecules from the surface, and in Figure 11 calibration curves of biosensor response to cortisol in human urine and saliva are shown. N values are a useful empirical measure of the biosensor performance. For urine, the N value for cortisol solution is 6.69 (as opposed to 2.85 for blank urine), and for saliva the N value is 6.51 (as opposed to 4.70 for blank saliva). N is closely related with the binding affinity of βCDs with cortisol; therefore, in the absence of cortisol N is lower.

Figure 11.

Figure 11

Open in a new tab

Capacitive determination of cortisol in multiple biofluids: (A) calibration results for determination of cortisol concentrations urine (N = 3); (B) calibration results for determination of cortisol concentrations in saliva (N = 3); (C) comparison between the biosensor response to addition of 2.5 nM cortisol solution vs addition of equivalent blank solutions; (D) comparison between the biosensor response to addition of 160 nM cortisol in PBS and 168 nM cortisol in urine and saliva solution vs equivalent blank solutions.

Urinary cortisol changes daily in the range of 11–281 nM,48 and the LOB and LOD for the surface were found to be 0.84 and 1.29 nM, respectively. In addition, salivary cortisol fluctuates in the range of 3.5–27 nM,49 and the LOB and LOD for our biosensor are 1.11 and 1.33 nM, respectively. Accordingly, it can be concluded that the proposed GC-carboxyphenyl-PPG:βCD biosensor can be used for detection of cortisol in urine and saliva samples.

3.7. Comparison of GC-carboxyphenyl-PPG:βCD with Other Cortisol Biosensors

We compared the results of GC-carboxyphenyl-PPG:βCD biosensor with other recently developed cortisol biosensors. Table 1 shows that while the LOD is comparable to that of other biosensors, GC-carboxyphenyl-PPG:βCD has three advantages:

  • Development of the nonenzymatic GC-carboxyphenyl-PPG:βCD biosensor is simple, and it does not require expensive and complicated procedures that are usually involved in development of immunoassays.

  • GC-carboxyphenyl-PPG:βCD is the first reusable cortisol biosensor, and it does not suffer from the common restrictions of immunoassays including low stability and low reproducibility. The surface is stable for at least a month, and it can be reused up to 10 times.

  • GC-carboxyphenyl-PPG:βCD can be used to detect cortisol in various relevant biological fluids such as urine (1.29–168 nM) and saliva (1.33–168 nM). Therefore, the developed biosensor is applicable to detect cortisol levels in urine or saliva samples of septic patients.

Table 1. Comparison of Various Recently Developed Biosensors for Detection of Cortisol.

biosensor detection method medium LOD (nM) reusability ref
Aptamer-based lateral flow assay Colorimetric Artificial saliva 2.73a Single use (4)
Human saliva 1.02a
Hydroxylamine-reduced silver nanoparticles (HA-AgNPs) Surface enhanced Raman spectroscopy (SERS) Bovine serum albumin in PBS 177b Single use (50)
Cortisol antibodies/suspension of magnetic beads Magnetically assisted SERS (MA-SERS) Human urine simulant and human serum 19.31b Single use (51)
Cortisol antibodies/gold-coated glass Rigid substrate SERS (RS-SERS) 8.28b
Cortisol antibody/PEDOT/Ni/Pt micromotor UV–visible spectrophotometry PBS buffer 275.89c Single use (52)
GC-carboxyphenyl-PPG:βCD electrochemical (EIS) PBS buffer 2.13a and 2.5c Reusable This work
Human urine 1.29a and 2.5c
Artificial saliva 1.33a and 2.5c
Open in a new tab
a

LOD based on eqs 8 and 9.

b

LOD, defined as 3 SDblank/s, where s is the slope of the calibration curve.

c

Lowest measured concentration.

4. Conclusion

We have designed a reusable biosensor platform that integrates the complexation of CDs with various hydrophobic guest molecules and polymers. In this platform, when a hydrophobic guest analyte is introduced into this system, the cyclodextrin will depart from the polymer support due to the competitive inclusion interaction between the analyte and cyclodextrin. In our previous work, we applied this technology to create an αCD based surface for detection of resveratrol. In this work, we first applied this technology to develop a cortisol biosensor based on competitive interactions of βCD with PPG and cortisol. Second, we improved the stability of biosensor by replacing gold–thiol bonds with stronger C–C triple bonds through diazonium salt chemistry. As a result of higher stability, the GC-carboxyphenyl-PPG:βCD could be regenerated and reused up to 10 times. Finally, the biosensor was employed for sensitive and selective sensing of cortisol in human urine and saliva.

Acknowledgments

This work was supported by NH EPSCoR NH BioMade Project which is provided by the National Science Foundation’s Research Infrastructure Improvement Award 1757371 and the Center for Integrated Biomedical and Bioengineering Research supported by Grant NIH P20 GM113131. The authors acknowledge the support of the College of Engineering and Physical Sciences and the Surface Enhanced Electrochemical Diagnostic Sensors (SEEDS) Laboratory at the University of New Hampshire.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsami.2c07701.

  • Quantification of Au 4f, C 1s, O 1s, and N 1s of gold electrode after different modification steps with XPS; quantification of C 1s, O 1s, and N 1s of carbon electrode after different modification steps with XPS; NMR studies on complex formation of βCD with PPG; NMR studies on complex formation of βCD with cortisol; NMR studies on complex formation of βCD with PPG and cortisol; stoichiometric ratios of βCD:PG and βCD:cortisol in the absence and presence of cortisol and PPG; high-resolution C 1s XPS spectra for carbon-carboxyphenyl-PPG:βCD surface before and after exposure to cortisol; high-resolution O 1s XPS spectra for carbon-carboxyphenyl-PPG:βCD surface before and after exposure to cortisol; high-resolution C 1s XPS spectra for carbon-carboxyphenyl-PPG:βCD surface before and after ethanol soak; high-resolution O 1s XPS spectra for carbon-carboxyphenyl-PPG:βCD surface before and after ethanol soak (PDF)

The authors declare no competing financial interest.

Supplementary Material

am2c07701_si_001.pdf (334.4KB, pdf)

References

  1. Tiao G.; Hobler S.; Wang J. J.; Meyer T. A.; Luchette F. A.; Fischer J. E.; Hasselgren P. O. Sepsis Is Associated with Increased MRNAs of the Ubiquitin-Proteasome Proteolytic Pathway in Human Skeletal Muscle. J. Clin. Invest. 1997, 99 (2), 163–168. 10.1172/JCI119143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Nguyen D. N.; Huyghens L.; Zhang H.; Schiettecatte J.; Smitz J.; Vincent J. L. Cortisol Is an Associated-Risk Factor of Brain Dysfunction in Patients with Severe Sepsis and Septic Shock. Biomed. Res. Int. 2014, 2014, 712742. 10.1155/2014/712742. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Holub M.; Džupová O.; Růžková M.; Stráníková A.; Bartáková E.; Máca J.; Beneš J.; Herwald H.; Beran O. Selected Biomarkers Correlate with the Origin and Severity of Sepsis. Mediators Inflamm. 2018, 2018, 1–11. 10.1155/2018/7028267. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Dalirirad S.; Han D.; Steckl A. J. Aptamer-Based Lateral Flow Biosensor for Rapid Detection of Salivary Cortisol. ACS Omega 2020, 5, 32890. 10.1021/acsomega.0c03223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Singh A.; Kaushik A.; Kumar R.; Nair M.; Bhansali S. Electrochemical Sensing of Cortisol: A Recent Update. Appl. Biochem. Biotechnol. 2014, 174 (3), 1115–1126. 10.1007/s12010-014-0894-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. vaz de Mello R. C.; Sad E. F.; Andrade B. C.; Neves S. P. F.; Santos S. M. E.; Sarquis M. M. S.; Marik P. E.; Dias E. P. Serum and Salivary Cortisol in the Diagnosis of Adrenal Insufficiency and as a Predictor of the Outcome in Patients with Severe Sepsis. Arq. Bras. Endocrinol. Metab. 2011, 55 (7), 455–459. 10.1590/S0004-27302011000700004. [DOI] [PubMed] [Google Scholar]
  7. Liu X.; Hsu S. P. C.; Liu W. C.; Wang Y. M.; Liu X.; Lo C. S.; Lin Y. C.; Nabilla S. C.; Li Z.; Hong Y.; Lin C.; Li Y.; Zhao G.; Chung R. J. Salivary Electrochemical Cortisol Biosensor Based on Tin Disulfide Nanoflakes. Nanoscale Res. Lett. 2019, 14, 189. 10.1186/s11671-019-3012-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Khan M. S.; Misra S. K.; Wang Z.; Daza E.; Schwartz-Duval A. S.; Kus J. M.; Pan D.; Pan D. Paper-Based Analytical Biosensor Chip Designed from Graphene-Nanoplatelet-Amphiphilic-Diblock-Co-Polymer Composite for Cortisol Detection in Human Saliva. Anal. Chem. 2017, 89, 2107. 10.1021/acs.analchem.6b04769. [DOI] [PubMed] [Google Scholar]
  9. Dhull N.; Kaur G.; Gupta V.; Tomar M. Highly Sensitive and Non-Invasive Electrochemical Immunosensor for Salivary Cortisol Detection. Sensors Actuators, B Chem. 2019, 293, 281–288. 10.1016/j.snb.2019.05.020. [DOI] [Google Scholar]
  10. Tuteja S. K.; Ormsby C.; Neethirajan S. Noninvasive Label-Free Detection of Cortisol and Lactate Using Graphene Embedded Screen-Printed Electrode. Nano-Micro Lett. 2018, 10 (3), 41. 10.1007/s40820-018-0193-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Sekar M.; Pandiaraj M.; Bhansali S.; Ponpandian N.; Viswanathan C. Carbon Fiber Based Electrochemical Sensor for Sweat Cortisol Measurement. Sci. Rep. 2019, 9 (1), 1–14. 10.1038/s41598-018-37243-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Zhou Q.; Kannan P.; Natarajan B.; Maiyalagan T.; Subramanian P.; Jiang Z.; Mao S. MnO2 Cacti-like Nanostructured Platform Powers the Enhanced Electrochemical Immunobiosensing of Cortisol. Sensors Actuators, B Chem. 2020, 317, 128134. 10.1016/j.snb.2020.128134. [DOI] [Google Scholar]
  13. Panahi Z.; Custer L.; Halpern J. M. Recent Advances in Non-Enzymatic Electrochemical Detection of Hydrophobic Metabolites in Biofluids. Sensors and Actuators Reports 2021, 3, 100051. 10.1016/j.snr.2021.100051. [DOI] [Google Scholar]
  14. Kim Y.; Yang J.; Hur H.; Oh S.; Lee H. H. Highly Sensitive Colorimetric Assay of Cortisol Using Cortisol Antibody and Aptamer Sandwich Assay. Biosensors 2021, 11 (5), 163. 10.3390/bios11050163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Leitão C.; Leal-Junior A.; Almeida A. R.; Pereira S. O.; Costa F. M.; Pinto J. L.; Marques C. Cortisol AuPd Plasmonic Unclad POF Biosensor. Biotechnol. Reports 2021, 29, e00587. 10.1016/j.btre.2021.e00587. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Briegel J.; Sprung C. L.; Annane D.; Singer M.; Keh D.; Moreno R.; Möhnle P.; Weiss Y.; Avidan A.; Brunkhorst F. M.; Fiedler F.; Vogeser M. Multicenter Comparison of Cortisol as Measured by Different Methods in Samples of Patients with Septic Shock. Intensive Care Med. 2009, 35 (12), 2151–2156. 10.1007/s00134-009-1627-9. [DOI] [PubMed] [Google Scholar]
  17. Gidwani B.; Vyas A. A Comprehensive Review on Cyclodextrin-Based Carriers for Delivery of Chemotherapeutic Cytotoxic Anticancer Drugs. Biomed Res. Int. 2015, 2015, 198268. 10.1155/2015/198268. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Tiwari G.; Tiwari R.; Rai A. Cyclodextrins in Delivery Systems: Applications. J. Pharm. Bioallied Sci. 2010, 2 (2), 72. 10.4103/0975-7406.67003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Panahi Z.; Merrill M. A.; Halpern J. M. Reusable Cyclodextrin-Based Electrochemical Platform for Detection of Trans -Resveratrol. ACS Appl. Polym. Mater. 2020, 2 (11), 5086–5093. 10.1021/acsapm.0c00866. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Filipović-Grčić J.; Voinovich D.; Moneghini M.; Bećirević-Laćan M.; Magarotto L.; Jalšenjak I. Chitosan Microspheres with Hydrocortisone and Hydrocortisone-Hydroxypropyl-β-Cyclodextrin Inclusion Complex. Eur. J. Pharm. Sci. 2000, 9 (4), 373–379. 10.1016/S0928-0987(99)00078-0. [DOI] [PubMed] [Google Scholar]
  21. Chang S. L.; Banga A. K. Transdermal Iontophoretic Delivery of Hydrocortisone from Cyclodextrin Solutions. J. Pharm. Pharmacol. 2011, 50 (6), 635–640. 10.1111/j.2042-7158.1998.tb06897.x. [DOI] [PubMed] [Google Scholar]
  22. Harada A.; Kamachi M. Complex Formation between Cyclodextrin and Poly(Propylene Glycol). J. Chem. Soc. Chem. Commun. 1990, 0 (19), 1322–1323. 10.1039/c39900001322. [DOI] [Google Scholar]
  23. Liu Y.; Yang Y. W.; Chen Y.; Zou H. X. Polyrotaxane with Cyclodextrins as Stoppers and Its Assembly Behavior. Macromolecules 2005, 38 (13), 5838–5840. 10.1021/ma047327v. [DOI] [Google Scholar]
  24. van der Meer S. B.; Seiler T.; Buchmann C.; Partalidou G.; Boden S.; Loza K.; Heggen M.; Linders J.; Prymak O.; Oliveira C. L. P.; Hartmann L.; Epple M. Controlling the Surface Functionalization of Ultrasmall Gold Nanoparticles by Sequence-Defined Macromolecules. Chem.—Eur. J. 2021, 27 (4), 1451–1464. 10.1002/chem.202003804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Díez-González S.; Marion N.; Nolan S. P. N-Heterocyclic Carbenes in Late Transition Metal Catalysis. Chem. Rev. 2009, 109 (8), 3612–3676. 10.1021/cr900074m. [DOI] [PubMed] [Google Scholar]
  26. Cao C.; Zhang Y.; Jiang C.; Qi M.; Liu G. Advances on Aryldiazonium Salt Chemistry Based Interfacial Fabrication for Sensing Applications. ACS Appl. Mater. Interfaces 2017, 9 (6), 5031–5049. 10.1021/acsami.6b16108. [DOI] [PubMed] [Google Scholar]
  27. Bollella P.; Hibino Y.; Kano K.; Gorton L.; Antiochia R. Enhanced Direct Electron Transfer of Fructose Dehydrogenase Rationally Immobilized on a 2-Aminoanthracene Diazonium Cation Grafted Single-Walled Carbon Nanotube Based Electrode. ACS Catal. 2018, 8 (11), 10279–10289. 10.1021/acscatal.8b02729. [DOI] [Google Scholar]
  28. Eissa S.; L’Hocine L.; Siaj M.; Zourob M. A Graphene-Based Label-Free Voltammetric Immunosensor for Sensitive Detection of the Egg Allergen Ovalbumin. Analyst 2013, 138 (15), 4378. 10.1039/c3an36883a. [DOI] [PubMed] [Google Scholar]
  29. Wang L.; Veselinovic M.; Yang L.; Geiss B. J.; Dandy D. S.; Chen T. A Sensitive DNA Capacitive Biosensor Using Interdigitated Electrodes. Biosens. Bioelectron. 2017, 87, 646–653. 10.1016/j.bios.2016.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. DuToit M.; Ngaboyamahina E.; Wiesner M. Pairing Electrochemical Impedance Spectroscopy with Conducting Membranes for the in Situ Characterization of Membrane Fouling. J. Membr. Sci. 2021, 618, 118680. 10.1016/j.memsci.2020.118680. [DOI] [Google Scholar]
  31. Margarit-Mattos I. C. P. EIS and Organic Coatings Performance: Revisiting Some Key Points. Electrochim. Acta 2020, 354, 136725. 10.1016/j.electacta.2020.136725. [DOI] [Google Scholar]
  32. Anandan V.; Gangadharan R.; Zhang G. Role of SAM Chain Length in Enhancing the Sensitivity of Nanopillar Modified Electrodes for Glucose Detection. Sensors 2009, 9 (3), 1295–1305. 10.3390/s90301295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sheng M.; Frurip D.; Gorman D. Reactive Chemical Hazards of Diazonium Salts. J. Loss Prev. Process Ind. 2015, 38, 114–118. 10.1016/j.jlp.2015.09.004. [DOI] [Google Scholar]
  34. Ribeiro J. A.; Sales M. G. F.; Pereira C. M. Electrochemistry-Assisted Surface Plasmon Resonance Biosensor for Detection of CA 15–3. Anal. Chem. 2021, 93, 7815. 10.1021/acs.analchem.0c05367. [DOI] [PubMed] [Google Scholar]
  35. Okada M.; Kawaguchi Y.; Okumura H.; Kamachi M.; Harada A. Complex Formation of Cyclodextrins with Poly (Propylene Glycol) Derivatives. J. Polym. Sci. Part A Polym. Chem. 2000, 38 (S1), 4839–4849. . [DOI] [Google Scholar]
  36. Shinzawa H.; Uchimaru T.; Mizukado J.; Kazarian S. G. Non-Equilibrium Behavior of Polyethylene Glycol (PEG)/Polypropylene Glycol (PPG) Mixture Studied by Fourier Transform Infrared (FTIR) Spectroscopy. Vib. Spectrosc. 2017, 88, 49–55. 10.1016/j.vibspec.2016.11.001. [DOI] [Google Scholar]
  37. Haziri V.; Phal S.; Boily J. F.; Berisha A.; Tesfalidet S. Oxygen Interactions with Covalently Grafted 2D Nanometric Carboxyphenyl Thin Films—An Experimental and DFT Study. Coatings 2022, 12 (1), 49. 10.3390/coatings12010049. [DOI] [Google Scholar]
  38. Orazem M. E.; Tribollet B.. Electrochemical Impedance Spectroscopy, 2nd ed.; John Wiley & Sons: Hoboken, NJ, 2017. [Google Scholar]
  39. Daniels J. S.; Pourmand N. Label-Free Impedance Biosensors: Opportunities and Challenges. Electroanalysis 2007, 19 (12), 1239–1257. 10.1002/elan.200603855. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Itagaki M.; Suzuki S.; Shitanda I.; Watanabe K. Electrochemical Impedance and Complex Capacitance to Interpret Electrochemical Capacitor. Electrochemistry 2007, 75 (8), 649–655. 10.5796/electrochemistry.75.649. [DOI] [Google Scholar]
  41. Thompson G.; Marnoto S.; Halpern J. M. Proper Controls to Electrochemically Evaluate Carotenoids Using β-Cyclodextrin Modified Surfaces. ECS Trans. 2017, 80 (10), 1177–1187. 10.1149/08010.1177ecst. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Ziino E.; Marnoto S.; Halpern J. M. Investigation to Minimize Electrochemical Impedance Spectroscopy Drift. ECS Trans. 2020, 97 (7), 737–745. 10.1149/09707.0737ecst. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Persoff P.; Thomas J. F. Estimating Michaelis-Menten or Langmuir Isotherm Constants by Weighted Nonlinear Least Squares. Soil Sci. Soc. Am. J. 1988, 52, 886. 10.2136/sssaj1988.03615995005200030052x. [DOI] [Google Scholar]
  44. Osmari T. A.; Gallon R.; Schwaab M.; Barbosa-Coutinho E.; Severo J. B.; Pinto J. C. Statistical Analysis of Linear and Non-Linear Regression for the Estimation of Adsorption Isotherm Parameters. Adsorpt. Sci. Technol. 2013, 31 (5), 433–458. 10.1260/0263-6174.31.5.433. [DOI] [Google Scholar]
  45. Huang Y.; Bell M. C.; Suni I. I. Impedance Biosensor for Peanut Protein Ara h 1. Anal. Chem. 2008, 80 (23), 9157–9161. 10.1021/ac801048g. [DOI] [PubMed] [Google Scholar]
  46. Armbruster D. A.; Pry T. Limit of Blank, Limit of Detection and Limit of Quantitation. Clin. Biochem. Rev. 2008, 29 (Suppl 1), S49–S52. [PMC free article] [PubMed] [Google Scholar]
  47. Wayu M. B.; DiPasquale L. T.; Schwarzmann M. A.; Gillespie S. D.; Leopold M. C. Electropolymerization of β-Cyclodextrin onto Multi-Walled Carbon Nanotube Composite Films for Enhanced Selective Detection of Uric Acid. J. Electroanal. Chem. 2016, 783, 192–200. 10.1016/j.jelechem.2016.11.021. [DOI] [Google Scholar]
  48. Lin C.-L.; Wu T.-J.; Machacek D. A.; Jiang N.-S.; Kao P. C. Urinary Free Cortisol and Cortisone Determined by High Performance Liquid Chromatography in the Diagnosis of Cushing’s Syndrome. J. Clin. Endocrinol. Metab. 1997, 82 (1), 151–155. 10.1210/jcem.82.1.3687. [DOI] [PubMed] [Google Scholar]
  49. Aardal E.; Holm A. C. Cortisol in Saliva — Reference Ranges and Relation to Cortisol in Serum. Clin. Chem. Lab. Med. 1995, 33 (12), 927–932. 10.1515/cclm.1995.33.12.927. [DOI] [PubMed] [Google Scholar]
  50. Moore T. J.; Sharma B. Direct Surface Enhanced Raman Spectroscopic Detection of Cortisol at Physiological Concentrations. Anal. Chem. 2020, 92 (2), 2052–2057. 10.1021/acs.analchem.9b04532. [DOI] [PubMed] [Google Scholar]
  51. Villa J. E. L.; Garcia I.; Jimenez de Aberasturi D.; Pavlov V.; Sotomayor M. D. P. T.; Liz-Marzán L. M. SERS-Based Immunoassay for Monitoring Cortisol-Related Disorders. Biosens. Bioelectron. 2020, 165, 112418. 10.1016/j.bios.2020.112418. [DOI] [PubMed] [Google Scholar]
  52. de Ávila B. E. F.; Zhao M.; Campuzano S.; Ricci F.; Pingarrón J. M.; Mascini M.; Wang J. Rapid Micromotor-Based Naked-Eye Immunoassay. Talanta 2017, 167, 651–657. 10.1016/j.talanta.2017.02.068. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

am2c07701_si_001.pdf (334.4KB, pdf)

Articles from ACS Applied Materials & Interfaces are provided here courtesy of American Chemical Society

RESOURCES