Академический Документы
Профессиональный Документы
Культура Документы
h i g h l i g h t s g r a p h i c a l a b s t r a c t
a r t i c l e i n f o a b s t r a c t
Article history: Urinary creatinine concentration is a critical physiological parameter that enables reliable assessment of
Received 30 October 2017 patient renal function and diagnosis of a broad spectrum of diseases. In this study, a simple and inex-
Accepted 6 December 2017 pensive sensor comprising monodisperse, citrate-capped silver nanoparticles (cc-AgNPs) was developed,
Available online 28 December 2017
which enabled rapid, sensitive and selective quantitation of creatinine directly in unprocessed urine. The
mechanism of this sensor entails the creatinine-mediated aggregation of the cc-AgNPs (within 1 min)
Keywords:
under alkaline conditions (pH 12). This is attributed to the tautomerization of creatinine to its amino
Citrate-capped silver nanoparticles
anionic species at alkaline pH, which cross-link the cc-AgNPs via hydrogen bond networks with the
Creatinine
Aggregation
negatively charged citrate caps. Creatinine elicited visibly-discernable color changes of the cc-AgNPs
Localized surface plasmon resonance (LSPR) colloids in a concentration-dependent manner up to 10 mM. UV-visible spectroscopic analyses of the
UV-Visible spectrophotometry cc-AgNPs revealed that creatinine elicited a concentration-dependent decrease in intensity of the
localized surface plasmon resonance (LSPR) band centered around 403 nm, with a concomitant increase
in intensity of the red-shifted LSPR band at 670 nm. This observation denotes a creatinine-mediated
* Corresponding author.
E-mail address: jonathan.blackburn@uct.ac.za (J.M. Blackburn).
1
Joint-first authors.
https://doi.org/10.1016/j.aca.2017.12.016
0003-2670/© 2018 Elsevier B.V. All rights reserved.
M.T. Alula et al. / Analytica Chimica Acta 1007 (2018) 40e49 41
increase in cc-AgNP particle size via aggregation, as confirmed by transmission electron microscopy
analysis. The cc-AgNP sensor exhibited a linear correlation between the A670/A403 extinction ratio and
creatinine concentration range of 0e4.2 mM in aqueous solutions (R2 ¼ 0.996), and a low detection limit
of 53.4 nM. Hence, the simplicity, short assay time, and high sensitivity and selectivity of our cc-AgNP
sensor affirms its utility as a creatinine monitoring assay for low-resource, point-of-care settings.
© 2018 Elsevier B.V. All rights reserved.
(TEM) images were obtained using a field emission transmission 3.2. Optimization of the creatinine detection assay
electron microscope (FE-TEM, JSM-6500F, JEOL Ltd., Tokyo, Japan),
operating at an accelerating voltage of 5 kV. The interaction of creatinine with AgNPs is highly dependent on
the pH of the reaction mixture. In the presence of sodium hy-
droxide, AgNPs interact preferentially with creatinine as opposed
2.3. Synthesis of citrate-capped silver nanoparticles
to other chemical species such as uric acid [14], which is likely due
to the pH-dependent tautomerization of creatinine. Here, the effect
Citrate-capped AgNPs were prepared based on the previously
of pH on the properties of cc-AgNP colloidal solutions, with appli-
reported method using ascorbic acid as the reducing agent and
cation to the detection of creatinine, was investigated systemati-
citrate as the stabilizer [30]. Briefly, 8.0 mL of aqueous solution
cally. Initially, three distinct mixtures of cc-AgNP colloids
containing a mixture of ascorbic acid (final concentration of
containing (a) NaOH only (12 mM NaOH final concentration, pH
0.6 mM) and trisodium citrate (final concentration of 3.0 mM) was
12); (b) 4.5 mM creatinine only; or (c) NaOH (12 mM NaOH final
prepared, and the pH adjusted to 10.5 with 0.1 M NaOH. To this
concentration, pH 12) plus 4.5 mM creatinine were prepared sepa-
solution, 80 mL of 0.1 M AgNO3 aqueous solution was added slowly
rately, and changes in color and LSPR bands of each were assessed
with gentle, continuous stirring at 30 C for 15 min. The color of the
after 1 min of incubation at room temperature (Fig. 1). In mixtures
solution changed from colorless to yellow immediately upon
(a) and (b), neither the color, nor the LSPR bands of the colloids
addition of AgNO3. The resulting yellow-colored, cc-AgNP colloids
changed relative to the pure cc-AgNPs colloid, which confirmed
were used to detect and quantitate creatinine following adjustment
that neither NaOH (12 mM, pH 12), nor creatinine alone (at pH 6.6),
of the pH of the colloids to 12 with 600 mM NaOH solution (final
respectively, could induce aggregation of the cc-AgNPs, in agree-
NaOH concentration: 12 mM).
ment with the study by Mohammadi et al. [29]. The absence of
aggregation of the cc-AgNPs in the presence of NaOH (12 mM, pH
2.4. General procedure for detection and quantitation of creatinine 12) alone can be explained by the negatively-charged state of cit-
rate resulting from complete deprotonation of all three carboxylic
Detection and quantitation of creatinine was performed at room groups at alkaline pH (pKa1: 3.13, pKa2: 4.76, pKa3: 6.40 for citrate),
temperature. Briefly, 10 mL of 600 mM NaOH solution was added thus ensuring strong electrostatic repulsion between the individual
into 480 mL of undiluted cc-AgNPs to adjust the pH of the colloids to cc-AgNPs. Moreover, El Badawy et al. specifically demonstrated that
12. Subsequently, 10 mL of various concentrations of creatinine were the zeta potential of cc-AgNPs decreases from 40 mV to ~-60 mV
added to the alkaline cc-AgNPs colloids. After incubating for 1 min, as the pH increases from 2 to 10, at ~10 mM NaCl ionic strength,
50 mL aliquots of the various creatinine/cc-AgNP colloid mixtures again without aggregation of the cc-AgNPs [32]. These observations
were diluted with deionized water to a final volume of 1 mL and confirm that Naþ and OH ions themselves do not induce aggre-
analyzed by UV-visible spectrophotometry. gation of the cc-AgNPs at a concentration of 12 mM.
In the third mixture (c), however, a significant color change of
the cc-AgNP colloids from yellow to dark blue was observed (Fig. 1,
2.5. Urine analysis
inset). In addition, the decrease in the LSPR band centered around
403 nm, with a concomitant red shift in the LSPR band to 670 nm
The detection and quantitation of creatinine in human urine was
were observed (Fig. 1, spectra). The decrease in the LSPR band
performed using a similar procedure to that employed for the
centered around 403 nm reflects a reduction in concentration of
detection of creatinine in aqueous solutions. Briefly, 15 mL of urine
individually-dispersed cc-AgNPs in suspension, while the red-shift
was sampled from a healthy male volunteer and diluted with 15 mL
in the LSPR band to 670 nm reflects an increase in cc-AgNP particle
deionized water. Thereafter, 10 mL of the diluted urine specimen
was mixed with 490 mL of alkaline cc-AgNP colloids. The urine/cc-
AgNP colloid mixture was incubated for 1 min at room tempera-
ture and analyzed by UV-visible spectrophotometry. 0.6
Fig. 2. TEM images of quasi-spherical cc-AgNPs (diameter: 27.4 ± 5.4 nm). (A) cc-AgNPs mixed with creatinine alone remained dispersed, whereas, (B) cc-AgNPs mixed with
creatinine and sodium hydroxide rapidly aggregated. The color inset images depict the corresponding colorimetric appearances of the bulk cc-AgNP colloids. (For interpretation of
the references to color in this figure legend, the reader is referred to the Web version of this article.)
size via aggregation, since the resonant wavelength of two coupled of reaction, and then remained almost constant over the remaining
cc-AgNPs significantly red-shifts relative to that of the individual, duration of observation. A similar trend was also reported by with
homogeneously-dispersed nanoparticles [33]. Importantly, the TDA-capped AgNPs in response to creatinine [29]. The sudden
degree of color change, and the decrease in LSPR intensity (around decrease in the LSPR band within the first 30 s of reaction, at higher
403 nm) of the colloids were highly dependent on the concentra- concentrations of creatinine, depicts the rapid association of the
tion of creatinine, as depicted in Fig. 4. It is known that nanoparticle homogenously dispersed cc-AgNPs into larger nanoparticle aggre-
size increases proportionally with the degree of aggregation. gates. Fig. 3B clearly illustrates this phenomenon with the extinc-
Therefore, the LSPR wavelength red-shifts with increasing particle tion spectra acquired from the mixture of cc-AgNPs and 10 mM
size due to the near-field coupling in the resonant wavelength peak creatinine at various time intervals up to 10 min. Since the aggre-
of the interacting nanoparticles. Collectively, these observations gation rate was fast, the difference in color changes at the different
suggest that, at alkaline pH, creatinine strongly interacts with, and time intervals beyond 1 min were difficult to identify colorimetri-
drives the aggregation of cc-AgNPs. cally (Fig. 3A). Additionally, extended reaction time may result in
The ability of creatinine to induce aggregation of cc-AgNPs was precipitation of the cc-AgNPs at higher creatinine concentrations
studied further by transmission electron microscopy (TEM). In the that, in turn, adversely affects the analytical linear range. Therefore,
presence of creatinine alone, the cc-AgNPs remained well dispersed 1 min was selected as the standard reaction time for subsequent
(Fig. 2A), which corresponded with the narrow LSPR band centered assay of creatinine both in aqueous solutions and human urine. The
around 403 nm (Fig. 1b). However, intense aggregation of the cc- rapid aggregation of the cc-AgNPs at alkaline pH, which occurred
AgNPs at alkaline pH occurred only upon addition of creatinine, only upon addition of creatinine, implies an alkaline-dependent,
as denoted by a rapid color change of the colloid (Fig. 2B), which chemical conformation change of creatinine that specifically
corresponds with the red-shift of the LSPR band to 670 nm (Fig. 1c). drives the aggregation of the cc-AgNPs. Importantly, Qin et al. [30]
These observations clearly demonstrate that trace amounts of demonstrated that the cc-AgNPs remained homogeneously
creatinine can induce the aggregation of cc-AgNPs, but, only at dispersed and exhibited progressive blue-shifting of the LSPR band
alkaline pH. Therefore, this approach obviates the need for the when incubated in increasingly alkaline pH at 100 C for 2 h, which
laborious surface modification of the AgNPs with molecules such as the authors attributed to an increase in sphericity of the cc-AgNPs
picric acid [28] and 2, 20 -thiodiacetic acid [29] to facilitate the due to Ostwald ripening. Furthermore, the method employed here
interaction between creatinine and bare AgNPs. to synthesize the cc-AgNPs results in complete consumption of Agþ
ions after 2 min of reaction at pH 10.5 [30]. Taking these factors into
3.3. Effect of reaction time on creatinine-mediated cc-AgNP consideration with our experimental conditions (10 min duration
aggregation at room temperature and pH 12), our observations are, therefore,
highly unlikely to be artefacts of nanoparticle instability issues, or
The effect of reaction time on the interaction between the cc- the reduction of unreacted Agþ ions to Ag0. Interestingly,
AgNPs and creatinine was investigated. Extinction ratios (A670/ Mohammadi et al. observed that their TDA-capped AgNPs exhibited
A403) were acquired for the cc-AgNPs incubated with different maximum absorbance in response to creatinine at pH 10, which
creatinine concentrations (0.5e10 mM) in alkaline medium (pH 12, decreased upon addition of either HCl or NaOH [29], whereas pH 12
12 mM NaOH final concentration) in 0.5 or 1 min time intervals, up was optimal for the detection of creatinine in our study. A plausible
to 10 min. Regardless of the concentration of creatinine, the ag- reason for the discrepancy between our observations and those of
gregation of cc-AgNPs plateaued, albeit to different degrees, within Mohammadi et al. is that citrate is a triprotic acid, whereas 2, 20 -
the first min of reaction i.e. after 1 min, no significant characteristic thiodiacetic acid is a diprotic acid, and therefore more sensitive to
changes in either spectral or colorimetric features were observed screening by Naþ ions or protonation by H3Oþ ions, which affects its
for the tested creatinine concentrations. As shown in Fig. 3A, this ability to hydrogen bond to creatinine.
effect is most pronounced at both 5 and 10 mM creatinine, whereby It is also important to consider the electrical double layer (EDL)
the A670/A403 extinction ratios increased sharply and linearly to theory, which states that the diffuse double layer of metal nano-
different, concentration-dependent intensities within the first 30 s particles decreases in thickness with increasing ionic strength,
44 M.T. Alula et al. / Analytica Chimica Acta 1007 (2018) 40e49
0.6 B 0 min
1.0 A 0.5 min
1 min
0.5 2 min
0.8
7 min
0.3 10 min
0.4
0.2
0.2 0.1
0.0 0.0
0 1 2 3 4 5 6 7 8 9 10 11 300 350 400 450 500 550 600 650 700 750 800
Incubation time (min) Wavelength (nm)
Fig. 3. (A) Effect of incubation time on A670/A403 extinction ratio of the cc-AgNP colloids incubated with creatinine concentrations of 0.5 (-) 1.0 ( ), 2.5 ( ), 5.0 ( ) and 10 mM ( ) in
alkaline medium (pH 12, 12 mM NaOH). (B) Extinction spectra of cc-AgNP colloids acquired after mixing and incubating with 10 mM creatinine in alkaline medium (pH 12, 12 mM
NaOH) over a time course of 0e10 min.
Fig. 4. (A) Image of cc-AgNP colloids mixed with different concentrations of creatinine at pH 12 (12 mM NaOH) depicting the colorimetric changes visible up to 10 mM creatinine, (B)
Changes in extinction spectra of cc-AgNP colloids in response to increasing concentrations of creatinine, up to 12 mM at pH 12 (12 mM NaOH), (C) Plot depicting the sigmoidal
relationship between the A670/A403 extinction ratio of the cc-AgNPs colloids and creatinine concentration at pH 12 (12 mM NaOH).
which, in turn, permits a greater degree of particle-particle inter- to the cc-AgNPs, was only 12 mM at most, and therefore insufficient
action resulting in aggregation of the nanoparticles. However, El alone to drive the aggregation of the cc-AgNPs.
Badawy et al. demonstrated that cc-AgNPs, with similar physico-
chemical properties to the nanoparticles prepared in this study, do 3.4. Assay sensitivity and detection limit
not aggregate at strong alkaline pH, due to a reduction in the zeta
potential of the cc-AgNPs, even in the presence of monovalent ions, The sensitivity of cc-AgNPs toward creatinine was determined
such as, Naþ, Cl and NO 3 , at 10 mM ionic strength [32]. Addi- by both colorimetry and UV-visible spectrophotometry using a
tionally, Huyhn and Chen demonstrated that divalent electrolytes reaction time of 1 min (Fig. 4AeC). Fig. 4A clearly illustrates the
such as, CaCl2 and MgCl2 were more efficient at inducing aggre- discernable colorimetric changes of the cc-AgNP colloids in a
gation of cc-AgNPs at significantly lower concentrations (2.1 mM creatinine concentration range of 0.5e10 mM. Fig. 4B depicts the
and 2.7 mM, respectively) than NaCl (47.6 mM) [34]. In the present spectral changes of the cc-AgNP colloids in response to increasing
experimental setup, divalent metal ions were absent, while the concentrations of creatinine after 1 min of incubation, where a
ionic strength conferred solely by Naþ ions, upon addition of NaOH gradual decrease in extinction intensity at 403 nm with a
M.T. Alula et al. / Analytica Chimica Acta 1007 (2018) 40e49 45
A670/A403
plateau in the A670/A403 extinction ratio above 8 mM creatinine
denotes saturation of the reaction due to the rapid and complete DF, ×5.33
aggregation of all homogeneously dispersed cc-AgNPs in suspen- 0.4
sion. A linear correlation was obtained between the A670/A403 DF, ×2.53
extinction ratio and creatinine concentration in the range of DF, ×1.66
0e4.2 mM, with a regression coefficient (R2) of 0.996 and the 0.2
DF, ×1.23
following equation;
DF, ×1.00
4 4 0.0
y ¼ 0.01655(±2.3347910 )x þ 0.05286(±2.9501710 ),
centration on the degree of nanoparticle aggregation induced by salt e none of which elicited changes in the A670/A403 extinction
creatinine and, in turn, its effect on the efficacy of quantitation of ratio relative to the blank (untreated) control. Therefore, it can also
creatinine. Different concentrations of cc-AgNP colloids were pre- be deduced that none of the inorganic anions, Cl, NO 2
3 , and SO4 ,
pared by diluting the synthesized colloids with ultrapure deionized act as interferents to induce cc-AgNPs aggregation at the tested
water to dilution factors (DF) of 1, 1.23, 1.66, 2.53, and 5.33. The concentrations. Only creatinine elicited a significant, ~24-fold in-
A670/A403 extinction ratios for each mixture comprising cc-AgNP crease in the A670/A403 extinction ratio, as well as a yellow-to-blue
colloids (at different dilution factors), and a series of creatinine
standard solutions were studied systematically. The changes in the
A670/A403 extinction ratios were used to assess the degree of cc-
AgNPs aggregation. We observed that the rate of creatinine-
1.0
mediated cc-AgNP aggregation was inversely proportional to the
concentration of the cc-AgNP colloids. Specifically, as the concen-
tration of the cc-AgNP colloids decreased, aggregation occurred 0.8
faster such that all homogeneously dispersed cc-AgNPs aggregated
completely, resulting in a very narrow analytical linear correlation
0.6
A670/A403
Glycine
Uric acid
Ascorbic acid
Creatinine
Glutamic acid
Glucose
Blank
colorimetric change of the cc-AgNPs colloid relative to the blank appearance and A670/A403 extinction ratio of the cc-AgNP colloids
control (Fig. 6). It is also important to note that only high purity were actually attributable to the spiked creatinine, over and above
creatinine (98% purity; Alfa Aesar) was used in this study, which the signal changes elicited by the endogenous creatinine of the
contained no contaminating ions that could potentially induce cc- urine sample. The duration of reaction, the changes in colorimetric
AgNPs aggregation. These observations are consistent with those appearance as well as the A670/A403 extinction ratio observed for
of Parmar et al., who tested the same set of inorganic and organic the cc-AgNP colloids mixed with creatinine-spiked, diluted urine
analytes investigated in this study, each at 10 mM, against picric samples, and diluted urine samples without spiked creatinine were
acid-capped AgNPs, but also observed selectivity of their sensor all similar to that of aqueous creatinine samples. Specifically, the
only toward creatinine [28]. Importantly, uric acid did not elicit any A670/A403 extinction ratio of the cc-AgNPs increased linearly with
change in the A670/A403 ratio compared to the blank control, due to the concentration of spiked creatinine in the diluted urine samples,
both uric acid and the cc-AgNPs being negatively-charged at alka- which was observed similarly for the pure aqueous creatinine so-
line pH, which ensures electrostatic charge repulsion between the lutions (Fig. 7A). Therefore, these observations suggest that the
two. This is a significant improvement on the study by Alula and components of the urinary matrix had been diluted sufficiently
Yang [14] in which the authors developed Ag@ZnO/Fe3O4 nano- such that they did not interfere with the interaction between
particle composites to detect creatinine in urine by SERS, but also creatinine and the cc-AgNPs under alkaline conditions.
observed significant interference from uric acid in the samples. An 8-point standard addition method was used to quantify
In summary, these observations confirm that the organic and endogenous creatinine in the urine sample (Fig. 7B). A linear cor-
inorganic analytes tested here do not interfere with our assay, due relation was obtained between the A670/A403 extinction ratio and
to their inability to induce cc-AgNP aggregation at equimolar con- the spiked creatinine concentration, with a regression coefficient
centrations, further validating the selectivity of the cc-AgNPs to- (R2) of 0.9803 and the following equation;
ward creatinine under alkaline conditions.
y ¼ 0.06841(±3.49103) x þ 0.06603(±3.48103),
3.6. Detection of creatinine in human urine where y is the A670/A403 extinction ratio, and x is the concentration
of spiked creatinine.
To demonstrate the practicality of our method, the sensor was Extrapolation of the linear correlation curve in Fig. 7B to the x-
further validated by detecting creatinine directly in a human urine intercept (y ¼ 0) yields a value of 0.965 ± 0.002 mM creatinine and
sample. The urine sample was initially tested without further therefore an actual urinary creatinine concentration of 965 ± 2 mM
pretreatment to remove the urinary matrix, which reportedly in- (0.965 ± 0.002 mM) in the tested sample (based on a 1000-fold
hibits aggregation of the nanoparticles [27]. In this assay, however, dilution of the sample), which is close to the lower end of the
the undiluted urine triggered intense aggregation of the cc-AgNPs, typical reported urine creatinine concentration range
such that rapid color change was observed, thereby impeding (1.77e26.52 mM) [35]. Thus, this result demonstrates the practical
quantitative analytical measurements. Hence, dilution of the urine application of our cc-AgNP sensor to the detection and quantitation
sample was necessary to reduce the concentration of creatinine and of creatinine in clinical urine samples with the necessary sensitivity
other matrix components present in the urine sample such that and specificity. The advantages and low limit of detection of our
changes in the A670/A403 extinction ratio and colorimetric appear- assay are highlighted in the context of the previously reported
ance of the cc-AgNPs fell within the linear regression range of the methods for the detection of creatinine in Table 1.
assay.
Different concentrations of creatinine were spiked into the
diluted urine sample, based on the standard addition method, and 3.7. Proposed mechanism of creatinine-mediated cc-AgNP
subsequently analyzed via the same protocol applied to the quan- aggregation at alkaline pH
titation of creatinine in pure aqueous solutions, as described earlier.
A diluted urine sample, without spiked creatinine, served as a Before describing the plausible mechanism by which creatinine
control to confirm that the additional changes in the colorimetric mediates aggregation of the cc-AgNPs at alkaline pH, it is necessary
A 0 μM
0.6 0.3 μM
0.5 μM 0.25 B
Extinction (arb. unit)
0.5 0.75 μM
1.0 μM
0.4 1.5 μM 0.20
A670/A403
2.0 μM
0.3 2.5 μM 0.15
3.0 μM
0.2 3.5 μM
4.0 μM 0.10
0.1
0.05
0.0
400 500 600 700 800 0.0 0.5 1.0 1.5 2.0 2.5
Wavelength (nm) Concentration of spiked creatinine (μM)
Fig. 7. (A) Extinction spectra of cc-AgNP colloids mixed with urine samples that were spiked with different concentrations of creatinine. The black spectrum corresponds to the
diluted urine sample without spiked pure creatinine (0 mM). (B) Calibration curve for the quantitation of creatinine in urine samples generated by the standard addition method.
M.T. Alula et al. / Analytica Chimica Acta 1007 (2018) 40e49 47
Table 1
Comparison of the present method with some previously reported methods employed in the detection of creatinine in urine.
1 High performance liquid chromatography 1.8 Good linear response achieved with an acceptable [3]
concentration range at physiologically-relevant levels.
2 Amperometric sensor 0.6 Based on the formation of copper-creatinine complex on a [6]
copper electrode surface. Uric acid was found to interfere
with the assay.
3 Capillary electrophoresis method 4.4 Laborious deproteination and centrifugation procedures [10]
were employed.
4 Gold nanoparticles-based colorimetric 80 The detection limit is high and procedure is costly. [25]
assay
5 Gold nanoparticles-based detection after 121.2 Laborious solid phase extraction was employed to [27]
solid phase extraction overcome urinary matrix interferences.
6 Simple, rapid, sensitive, and selective 0.066 The method requires no additional laborious surface Present work
sensor based on citrate-capped silver modifications of silver nanoparticles or conjugation with
nanoparticles other linker molecules, and yielded a wide linear correlation
between the cc-AgNPs extinction ratio and concentration of
creatinine.
to briefly review the nature of the surface interaction between findings by Antoniou et al. The amino tautomer, which is the most
citrate molecules and the AgNPs. Frost et al. demonstrated, by thermodynamically stable form of creatinine [40], is therefore
attenuated total reflectance Fourier transform infrared (ATR-FTIR) favored at alkaline pH via formation of the carbanion and oxoanion
spectroscopy, that two carboxylate groups as well as the hydroxyl species, with the negative charge localized over the C5 or O8 atoms,
group of citrate bind to the surface of AgNPs (Fig. 8) [36]. Since two respectively (Fig. 8). Thus, stable hydrogen bonds likely form be-
of the three carboxylate groups of each surface-interacting citrate tween the carbanion/oxoanion amino tautomers and the single,
molecule are bound to the surface of the AgNP, only the remaining exposed negatively-charged carboxylate group of the citrate caps.
carboxylate group is available to contribute to the negative surface Furthermore, Naþ ions can electrostatically interact with the
charge of the AgNP. At mildly acidic conditions (pH 6.6 in this carbanion/oxoanion species of the surface bound creatinine tau-
study), all carboxylate groups of citrate are fully deprotonated tomers, which further screen the negatively-charged citrate caps of
(pKa1: 3.13, pKa2: 4.76, pKa3: 6.40), and remain fully deprotonated at the cc-AgNPs from each other. As a result, an intermolecular
alkaline conditions (pH 12 in this study) (Fig. 8), which sustains the hydrogen bonding network would form between surface-bound
electrostatic repulsion between the individual cc-AgNPs, thus creatinine molecules that effectively ‘cross-link’ the interacting
maintaining their stabilization in a homogeneously dispersed state cc-AgNPs, thus leading to aggregation. It should also be mentioned
[32]. that Gangopadhyay et al. recently demonstrated, by SERS-based
Creatinine has a pKa1 of 4.9 and pKa2 of 9.2 [37]. At pH values reaction monitoring, the decyclization of creatinine to creatine at
close to its pKa1, creatinine rapidly interconverts between its two pH 12; however, this was observed only after 30 min of reaction
neutral tautomers i.e. 2-amino-1-methyl-imidazoline-4-one [41]. Given the reaction time of 1 min at pH 12 in our study,
(amino tautomer) and 2-imino-1-methyl-imidazolidine-4-one creatinine decyclization is therefore not a likely cause of the
(imino tautomer) [38], with both tautomers capable of inter- creatinine-mediated aggregation of the cc-AgNPs.
converting with the creatininium cation as well, as observed by
Raman spectroscopy [39] (Fig. 8). Furthermore, Gao et al. demon-
4. Conclusions
strated that creatinine also exists in equilibrium between its amino,
imino, and creatininium cation tautomers at pH 6 [39]. In the
The detection and quantitation of creatinine in urine specimens
present study, we observed that creatinine failed to induce aggre-
is an important test that enables clinicians to diagnose renal
gation of the cc-AgNPs at pH 6.6, where the amino, imino, and
dysfunction occurring in a broad spectrum of diseases. The ability
creatininium cation tautomers are still likely at equilibrium with
to rapidly and reliably quantitate creatinine with high sensitivity
each other. Although the amino, imino and creatininium cation
and specificity in a complex specimen matrix such as urine, without
tautomers can form hydrogen bonds with the single exposed
sophisticated analytical equipment, is especially useful in low-
carboxylate group of each surface-adsorbed citrate molecule, these
resource settings. To address this need, we developed a simple
intermolecular hydrogen bonds are likely to be only transient due
and inexpensive, colorimetric citrate-capped AgNP sensor, which
to the rapid interconversion between these tautomeric forms of
demonstrated sensitive and selective detection of creatinine in
creatinine (Fig. 8).
aqueous solutions and in human urine within one minute. The
Above neutral pH, however, Antoniou et al. observed by mass
detection mechanism relies on the creatinine-mediated aggrega-
spectrometry that the anionic species of creatinine became
tion of the cc-AgNPs at pH 12, which elicits a visibly-detectable
increasingly prevalent as the pH increased, which reached a
color change. The ccAgNP sensor exhibited a linear correlation
maximum ion count (~3290) at pH 11.1, followed by a sharp
between the A670/A403 extinction ratio and creatinine concentra-
decrease in ion count at higher pH (~500 at pH 12.5) [37]. Fig. 8
tion range of 0e4.2 mM (R2 ¼ 0.996), and a low detection limit of
(adapted from Antoniou et al.) depicts the formation of the carb-
53.4 nM. Importantly, this approach requires no additional surface
anion and oxoanion species exclusively from the amino tautomer of
modification of the AgNPs, or conjugation with other linker mole-
creatinine at alkaline pH [37]. At pH 12 e at which we observed
cules that would render synthesis of the AgNPs laborious and
creatinine-mediated aggregation of the cc-AgNPs e the carbanion
detection complicated. Furthermore, neither sample pretreatment,
and oxoanion species of the amino tautomer would still be present,
nor selective extraction of creatinine are required to facilitate its
albeit at ~46% of the maximum level at pH 11.1, according to the
detection in urine with our method. The advantages of simplicity,
48 M.T. Alula et al. / Analytica Chimica Acta 1007 (2018) 40e49
Fig. 8. Schematic diagram depicting the pH-dependent tautomerism of creatinine, and its bonding interactions with citrate-capped AgNPs under mild acidic, and alkaline con-
ditions. Aggregation of the cc-AgNPs, at pH 12, is mediated by formation of a hydrogen bonding network between the amine group of the carbanion/oxoanion amino tautomers of
creatinine and the negatively-charged, carboxylate group of the surface-adsorbed citrate molecules. Additionally, electrostatic interactions between Naþ ions and the carbanion/
oxoanion amino tautomers of creatinine further screen the negatively-charged citrate groups between individual cc-AgNPs in close proximity.
short assay duration, high sensitivity and selectivity demonstrate Partnership. LK thanks the NRF (Grant No. 99517) for a postdoctoral
the utility of our cc-AgNP sensor as a creatinine monitoring assay innovation fellowship; MTA and NRH were supported by post-
for low-resource, point-of-care settings. doctoral fellowships from the SAMRC.
Acknowledgements References
JMB thanks the National Research Foundation (NRF) for a South [1] R.D. Perrone, N.E. Madias, A.S. Levey, Serum creatinine as an index of renal-
African Research Chair grant (Grant No. 64760). This research was function - new insights into old concepts, Clin. Chem. 38 (1992) 1933e1953.
[2] S. Hallan, A. Åsberg, M. Lindberg, H. Johnsen, Validation of the modification of
supported by a grant to JMB from the South African Medical diet in renal disease formula for estimating GFR with special emphasis on
Research Council's (SAMRC) Strategic Health Innovation calibration of the serum creatinine assay, Am. J. Kidney Dis. 44 (2004) 84e93.
M.T. Alula et al. / Analytica Chimica Acta 1007 (2018) 40e49 49
[3] J. Zhao, H. Chen, P. Ni, B. Xu, X. Luo, Y. Zhan, P. Gao, D. Zhu, Simultaneous based on its effect on the growth of gold nanoparticles induced by the reac-
determination of urinary tryptophan, tryptophan-related metabolites and tion of chloroauric acid and hydroxylamine, Microchim. Acta 183 (2016)
creatinine by high performance liquid chromatography with ultraviolet and 3135e3141.
fluorimetric detection, J. Chromatogr. B 879 (2011) 2720e2725. [23] A. Safavi, R. Ahmadi, Z. Mohammadpour, Colorimetric sensing of silver ion
[4] A.K. Hewavitharana, H.L. Bruce, Simultaneous determination of creatinine and based on anti-aggregation of gold nanoparticles, Sensor. Actuator. B Chem.
pseudouridine concentrations in bovine plasma by reversed-phase liquid 242 (2017) 609e615.
chromatography with photodiode array detection, J. Chromatogr. B 784 [24] Y. Zhao, L. Gui, Z. Chen, Colorimetric detection of Hg2þ based on target-
(2003) 275e281. mediated growth of gold nanoparticles, Sensor. Actuator. B Chem. 241 (2017)
[5] C.J. Kochansky, T.G. Strein, Determination of uremic toxins in biofluids: 262e267.
creatinine, creatine, uric acid and xanthenes, J. Chromatogr. B 747 (2000) [25] Y. He, X. Zhang, H. Yu, Gold nanoparticles-based colorimetric and visual
217e227. creatinine assay, Microchim. Acta 182 (2015) 2037e2043.
[6] C.-H. Chen, M.S. Lin, A novel structural specific creatinine sensing scheme for [26] J. Du, B. Zhu, W.R. Leow, S. Chen, T.C. Sum, X. Peng, X. Chen, Colorimetric
the determination of the urine creatinine, Biosens. Bioelectron. 31 (2012) detection of creatinine based on plasmonic nanoparticles via synergistic co-
90e94. ordination chemistry, Small 11 (2015) 4104e4110.
[7] W.R. de Araújo, M.O. Salles, T.R.L.C. Paixa ~o, Development of an enzymeless [27] J. Sittiwong, F. Unob, Detection of urinary creatinine using gold nanoparticles
electroanalytical method for the indirect detection of creatinine in urine after solid phase extraction, Spectrochim. Acta 138 (2015) 381e386.
samples, Sensor. Actuator. B Chem. 173 (2012) 847e851. [28] A.K. Parmar, N.N. Valand, K.B. Solanki, S.K. Menon, Picric acid capped silver
[8] J.-C. Chen, A.S. Kumar, H.-H. Chung, S.-H. Chien, M.-C. Kuo, J.-M. Zen, An nanoparticles as a probe for colorimetric sensing of creatinine in human blood
enzymeless electrochemical sensor for the selective determination of creati- and cerebrospinal fluid samples, Analyst 141 (2016) 1488e1498.
nine in human urine, Sensor. Actuator. B Chem. 115 (2006) 473e480. [29] S. Mohammadi, G. Khayatian, Highly selective and sensitive photometric
[9] J.A. Mun ~ oz, M. Lopez-Mesas, M. Valiente, Development and validation of a creatinine assay using silver nanoparticles, Microchim Acta 182 (2015)
simple determination of urine metabolites (oxalate,citrate, uric acid and 1379e1386.
creatinine) by capillary zone electrophoresis, Talanta 81 (2010) 392e397. [30] Y. Qin, X. Ji, J. Jing, H. Liu, H. Wu, W. Yang, Size control over spherical silver
[10] A.C.O. Costa, J.L. da Costa, F.G. Tonin, M.F.M. Tavares, G.A. Micke, Development nanoparticles by ascorbic acid reduction, Colloid. Surface. 372 (2010)
of a fast capillary electrophoresis method for determination of creatinine in 172e176.
urine samples, J. Chromatogr. A 1171 (2007) 140e143. [31] L. Malassis, R. Dreyfus, R.J. Murphy, L.A. Hough, B. Donnio, C.B. Murray, One-
[11] R. Harlan, W. Clarke, J.M. Di Bussolo, M. Kozak, J. Straseski, D.L. Meany, An step green synthesis of gold and silver nanoparticles with ascorbic acid and
automated turbulent flow liquid chromatographyeisotope dilution mass their versatile surface post-functionalization, RSC Adv 6 (2016) 33092e33100.
spectrometry (LCeIDMS) method for quantitation of serum creatinine, Clin. [32] A.M. El Badawy, T.P. Luxton, R.G. Silva, K.G. Scheckel, M.T. Suidan,
Chim. Acta 411 (2010) 728e1734. T.M. Tolaymat, Impact of environmental conditions (pH, ionic strength, and
[12] S. Chen, L. Chen, J. Wang, J. Hou, Q. He, J. Liu, J. Wang, S. Xiong, G. Yang, Z. Nie, electrolyte type) on the surface charge and aggregation of silver nanoparticles
2, 3, 4, 5-Tetrakis (30 , 40 -dihydroxylphenyl) thiophene: a new matrix for the suspensions, Environ. Sci. Technol. 44 (2010) 1260e1266.
selective analysis of low molecular weight amines and direct determination of [33] K.-H. Su, Q.-H. Wei, X. Zhang, J.J. Mock, D.R. Smith, S. Schultz, Interparticle
creatinine in urine by MALDI-TOF MS, Anal. Chem. 84 (2012) 10291e10297. coupling effects on plasmon resonances of nanogold particles, Nano Lett. 3
[13] K. Huang, M. Li, H. Li, M. Li, Y. Jiang, X. Fang, Accurate quantification of (2003) 1087e1090.
creatinine in serum by coupling a measurement standard to extractive elec- [34] K.A. Huynh, K.L. Chen, Aggregation kinetics of citrate and poly-
trospray ionization mass spectrometry, Sci. Rep. 6 (2016) 19283. vinylpyrrolidone coated silver nanoparticles in monovalent and divalent
[14] M.T. Alula, J. Yang, Photochemical decoration of magnetic composites with electrolyte solutions, Environ. Sci. Technol. 45 (2011) 5564e5571.
silver nanostructures for determination of creatinine in urine by surface- [35] D.B. Barr, L.C. Wilder, S.P. Caudill, A.J. Gonzalez, L.L. Needham, J.L. Pirkle,
enhanced Raman spectroscopy, Talanta 130 (2014) 55e62. Urinary creatinine concentrations in the U.S. population: implications for
[15] M. Li, Y. Du, F. Zhao, J. Zeng, C. Mohan, W.-C. Shih, Reagent- and separation- urinary biologic monitoring measurements, Environ. Health Perspect. 113
free measurements of urine creatinine concentration using stamping surface (2005) 192e200.
enhanced Raman scattering (S-SERS), Biomed. Opt. Express 6 (2015) 849. [36] M.S. Frost, M.J. Dempsey, D.E. Whitehead, The response of citrate function-
[16] P. Campins Falco , L.T. Genaro, S.M. Lloret, F.B. Gomez, A.S. Cabeza, C.M. Legua, alised gold and silver nanoparticles to the addition of heavy metal ions,
Creatinine determination in urine samples by batchwise kinetic procedure Colloid. Surface. Physicochem. Eng. Aspect. 518 (2017) 15e24.
and flow injection analysis using the Jaffe reaction: chemometric study, [37] M.G. Antoniou, D.D. Dionysiou, Application of immobilized titanium dioxide
Talanta 55 (2001) 1079e1089. photocatalysts for the degradation of creatinine and phenol, model organic
[17] S. Link, M.A. El-Sayed, Optical properties and ultrafast dynamics of metallic contaminants found in NASA's spacecrafts wastewater streams, Catal. Today
Nanocrystals, Annu. Rev. Phys. Chem. 54 (2003) 331e366. 124 (2007) 215e223.
[18] Y. Song, W. Wei, X. Qu, Colorimetric biosensing using smart materials, Adv. [38] D. Kotsyubynskyy, S. Molchanov, A. Gryff-Keller, Creatinine and creatininium
Mater. 23 (2011) 4215e4236. cation in DMSO-d6 solution. Structure and restricted internal rotation of NH2
[19] Y. Li, Q. Wang, X. Zhou, C.-Y. Wen, J. Yu, X. Han, X. Li, Z.-F. Yan, J. Zeng, group, Magn. Reson. Chem. 42 (2004) 1027e1036.
A convenient colorimetric method for sensitive and specific detection of cy- [39] J. Gao, Y. Hu, S. Li, Y. Zhang, X. Chen, Tautomeric Equilibrium of Creatinine and
anide using Ag@Au coreeshell nanoparticles, Sensor. Actuator. B Chem. 228 Creatininium Cation in Aqueous Solutions Explored by Raman Spectroscopy
(2016) 366e372. and Density Functional Theory Calculations, vol 410, 2013, pp. 81e89.
[20] M. Rycenga, C.M. Cobley, J. Zeng, W. Li, C.H. Moran, Q. Zhang, D. Qin, Y. Xia, [40] A.R. Butler, C. Glidewell, Creatinine: an examination of its structure and some
Controlling the synthesis and assembly of silver nanostructures for plasmonic of its reactions by synergistic use of MNDO calculation and nuclear magnetic
applications, Chem. Rev. 111 (2011) 3669e3712. resonance spectroscopy, J. Chem. Soc. Perkin Trans. II 9 (1985) 1465e1467.
[21] L. Polavarapu, J. Pe rez-Juste, Q.-H. Xu, L.M. Liz-Marz an, Optical sensing of [41] D. Gangopadhyay, P. Sharma, S.K. Singh, P. Singh, V. Deckert, J. Popp,
biological, chemical and ionic species through aggregation of plasmonic R.K. Singh, Surface enhance Raman scattering based reaction monitoring of
nanoparticles, J. Mater. Chem. C 2 (2014) 7460e7476. in vitro decyclization of creatinine / creatine, RSC Adv. 6 (2016)
[22] S. Wang, X. Hu, L. Tan, Q. Liao, Z. Chen, Colorimetric detection of lysozyme 58943e58949.