Reaction of gold nanoparticles with tetracyanoquinoidal molecules. Spectrophotometric determination of the Au0 content of gold nanoparticles.

Abstract

The interaction of gold nanoparticles (AuNPs) and typical tetracyanoquinoidal compounds such as bis(dicyanomethylene)-bithiophene and tetracyanoquinodimethane (TCNQ) has been investigated. AuNPs in toluene solution reduce the tetracyano compounds to the radical anion, as shown by UV-vis spectroscopy. The reaction, promoted by the bromide anion used as a stabilizer for AuNPs, involves in the case of TCNQ the total amount of Au(0) in the nanoparticles. A spectrophotometric method for the evaluation of the Au(0) content of capped AuNPs in organic solution has been established and successfully applied to the analysis of dodecanethiol-capped AuNPs.

Full text

Page 1

Reaction of Gold Nanoparticles with
Tetracyanoquinoidal Molecules.
Spectrophotometric Determination of the Au(0)
Content of Gold Nanoparticles
Gianni Zotti* and Barbara Vercelli
Istituto CNR per l’Energetica e le Interfasi, c.o Stati Uniti 4, 35127 Padova, Italy
Anna Berlin
Istituto CNR di Scienze e Tecnologie Molecolari via C.Golgi 19, 20133 Milano, Italy
The interaction of gold nanoparticles (AuNPs) and typical
tetracyanoquinoidal compounds such as bis(dicyanom-
ethylene)-bithiophene and tetracyanoquinodimethane
(TCNQ) has been investigated. AuNPs in toluene solution
reduce the tetracyano compounds to the radical anion,
as shown by UV-vis spectroscopy. The reaction, pro-
moted by the bromide anion used as a stabilizer for
AuNPs, involves in the case of TCNQ the total amount of
Au(0) in the nanoparticles. A spectrophotometric method
for the evaluation of the Au(0) content of capped AuNPs
in organic solution has been established and successfully
applied to the analysis of dodecanethiol-capped AuNPs.
INTRODUCTION
Gold nanoparticles (AuNPs) are particularly investigated due
to their potential applications in sensors, nanoelectronic devices,
biochemical reagents, and catalysts.1We are interested in their
connection through conjugated pathways, such as those provided
by oligothiophenes and, in general, conjugated molecules.2The
possible use of such structures in optoelectronic devices (such
as solar cells, light emitting diodes, electronic memories, etc.) is
being explored intensively.3
In search of novel properties of AuNPs by interaction with
organic semiconductors such as bis(dicyanomethylene)-oligoth-
iophenes,4we have investigated their reciprocal interaction.
Here, we report on the reaction of 5,5′-bis(dicyanomethylene)-
5,5′-dihydro-2,2′-bithiophene(T2CN4)andtetracyanoquin-
odimethane (TCNQ), illustrated in Chart 1, with tetraoctylammo-
nium bromide-stabilized AuNPs in toluene, which is its most
common form in organic medium.5The investigation, performed
with the use of UV-vis spectroscopy and electrochemistry, has
shown that TCNQ operates the complete oxidation of the zerova-
lent gold content of the nanoparticles, thus allowing the establish-
ment of a novel, easy, and fast method for the analysis of gold
nanoclusters in organic solvents.
EXPERIMENTAL SECTION
Chemicals and Reagents. T2CN4 was prepared as reported
in the literature.6TCNQ (98%, from Fluka) was recrystallized from
acetonitrile, which was reagent grade (Uvasol, Merck) with a
water content of <0.01%. The supporting electrolyte tetrabuty-
lammonium perchlorate (Bu4NClO4) and all other chemicals were
reagent grade and used as received.
Toluene solutions of gold nanoparticles (AuNPs, 10-2M or 2
g L-1gold concentration by ICPMS, λmax) 525 nm), stabilized
by 0.05 M tetraoctylammonium bromide (TOABr), were prepared
according to Schiffrin.5From TEM analysis, the average particle
size of the Au clusters is 5 ( 1 nm.
Dodecanethiol-capped gold nanoparticles (C12-Au) were pre-
pared according to the procedure of Brust.7
Gold electrodes were 1 × 4 cm2sheets that were treated for
1 min with hot mixed chromic acid (K2Cr2O7in 96% H2SO4), then
carefully washed with Milli-Q water and dried.
Apparatus and Procedure. Spectrophotometric experiments
were performed at room temperature in 0.2-cm-path quartz cells
with nitrogen-degassed solutions. Caps and parts in possible
contact with the solutions were in Teflon to avoid accidental
interferences. Spectra were obtained with a Perkin-Elmer Lambda
15 spectrometer.
Electrochemical experiments were performed at room tem-
perature under nitrogen in three-electrode cells. The counter
* To whom correspondence should be addressed. Phone: (39) 049-829-5868.
Fax: (39) 049-829-5853. E-mail: g.zotti@ieni.cnr.it.
(1) Daniel, M. C.; Astruc, D. Chem. Rev. 2004, 104, 293.
(2) Zotti, G.; Vercelli, B.; Berlin, A.; Battagliarin, M.; Herna ´ndez, V.; Lo ´pez
Navarrete, J. T. J. Phys. Chem. C 2007, 111, 5886.
(3) (a) Huynh, W. U.; Dittmer, J. J.; Alivisatos, A. P. Science 2002, 295, 2425.
(b) Dabbousi, B. O.; Bawendi, M. G.; Onitsuka, O.; Rubner, M. F. Appl.
Phys. Lett. 1995, 66, 1316. (c) Tseng, R. J.; Huang, J.; Ouyang, J.; Kaner,
R. B.; Yang, Y. Nano Lett. 2005, 5, 1077. (d) Franke, M. E.; Koplin, T. J.;
Simon, U. Small 2006, 2, 36.
(4) Janzen, D. E.; Burand, M. W.; Ewbank, P. C.; Pappenfus, T. M.; Higuchi,
H.; da Silva Filho, D. A.; Young, V. G.; Bredas, J. L.; Mann, K. R. J. Am.
Chem. Soc. 2004, 126, 15295.
(5) Brust, M.; Bethell, D.; Schiffrin, D. J.; Kiely, C. Adv. Mater. 1995, 7, 795.
(6) Hernandez, V.; Hotta, S.; Lopez Navarrete, J. T. J. Chem. Phys. 1998, 109,
2543.
(7) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. J. J. Chem.
Soc., Chem. Commun. 1994, 801.
Anal. Chem. 2008, 80, 815-818
10.1021/ac701715y CCC: $40.75
Published on Web 01/10/2008
© 2008 American Chemical Society
Analytical Chemistry, Vol. 80, No. 3, February 1, 2008
815

Page 2

electrode was platinum; the reference electrode was silver/0.1 M
silver perchlorate in acetonitrile (0.34 V vs SCE). The voltammetric
apparatus (AMEL, Italy) included a 551 potentiostat modulated
by a 568 programmable function generator and coupled to a 731
digital integrator. The working electrode for cyclic voltammetry
was a platinum or gold minidisc electrode (0.003 cm2), depending
on the use.
RESULTS AND DISCUSSION
The optical and electrochemical parameters of the tetracyano-
quinoidal molecules here investigated and of their radical anion
forms, which are of interest for the present study, are summarized
in Table 1. The UV-vis spectrum of the neutral quinoidal
compounds is dominated by a strong absorption attributable to
the HOMO-LUMO electronic transition. In their radical anion
forms, the bands are strongly red-shifted and become structured.
Considering the electrochemical properties, the compounds are
reversibly reduced in two one-electron processes.
AuNPs and T2CN4. The strong absorption in the UV-vis
spectrum of T2CN4 in toluene at 543 nm, which gives the sample
a fuchsia color, turns to a light blue with a maximum absorption
at 651 nm upon addition of AuNPs (Figure 1a).
The same occurs by addition to the T2CN4 solution of
triethylamine or by one-electron electrochemical reduction. The
band of the neutral molecule is, in fact, batochromically shifted
in the radical anion, which is similar to the behavior of T3CN4.8
Thus, the reaction with gold appears to be simply a reduction.
The differences are that gold acts faster then the amine, but
nothing is known about the extent of its reaction.
AuNPs and TCNQ. AuNPs behave similarly toward TCNQ.
Analysis of the spectral evolution of AuNPs with addition of TCNQ
(Figure 1b) shows the appearance of strong vibronically structured
bands at 423, 753, and 856 nm, attributable to the radical anion of
TCNQ.9Thus, it appears that, similar to T2CN4, AuNPs act as
reductants vs TCNQ.
This result could be a residual effect of the reducing agent
used in the synthesis. In fact, rest potentials of solutions of as-
prepared AuNPs (e.g., Au140(C6)53) are generally more negative
(ca. -0.2 V vs Ag/Ag+) than the potential of zero charge of the
AuNPs, indicating that the reduction step in synthesis leaves a
residue of reductive charge on the particles.10
Another possibility is that reduction is operated by some
disaggregated and therefore more reactive gold atoms. It has, in
fact, been recently reported11that ligand exchange reactions using
multivalent coordinating polymers such as polyethylenimine are
able to partially disaggregate preformed colloidal gold nanocrys-
tals, leading to atomic gold clusters consisting of only eight gold
atoms. About 30% of the Au atoms of the nanocrystals were in
that case disaggregated bu,t in fact, converted into smaller atomic
clusters.
Both the suggested hypotheses require the partial involvement
of the gold atoms. It therefore appeared that a quantitative
determination of the reducing power of AuNPs was required.
To this end, we have performed a spectrophotometric titration
of TCNQ reduction to radical anion by added AuNPs. The fast
development of the band at 856 nm has indicated a clear equivalent
point corresponding to one exchanged electron (or TCNQ anion)
per gold atom. This result is the same from a typically prepared
AuNPs batch and one that was further treated with sulfuric acid,
which indicates that no residual sodium borohydride was left in
the gold solution. We must here say that commercial TOABr
partially reduces TCNQ via an impurity (probably residual
trioctylamine), but its contribution to the titration did not exceed
1% in our tests.
Treatment of a AuNPs solution (typically 10-3M) in toluene
with 1:1 TCNQ causes, in addition to the formation of the TCNQ
radical anion described above, the complete disappearance of the
gold SP band.
The previous results indicate that the reducing charge largely
exceeds that expected from partially disaggregated gold atoms
or negative charges stored in the large gold nanoparticles; namely,
(8) Pappenfus, T. M.; Raff, J. D.; Hukkanen, E. J.; Burney, J. R.; Casado, J.;
Drew, S. M.; Miller, L. L.; Mann, K. R. J. Org. Chem. 2002, 67, 6015.
(9) Melby, L. R.; Harder, R. J.; Hertler, W. R.; Mahler, W.; Benson, R. E.; Mochel,
W. E. J. Am. Chem. Soc. 1962, 84, 3374.
(10) Wuelfing, W. P.; Green, S. J.; Pietron, J. J.; Cliffel, D. E.; Murray, R. W. J.
Am. Chem. Soc. 2000, 122, 11465.
(11) Duan, H.; Nie, S. J. Am. Chem. Soc. 2007, 129, 2412.
Chart 1
Table 1. Reduction Potentials (Ered
Tetracyano Compounds and Maximum Absorption
Wavelength in CH2Cl2of their Neutral (λ) and Radical
Anion (λ(-)) Forms
0) vs Ag/Ag+of
compound
T2CN4
TCNQ
TCNE
Ered
0/V
λ/nm
550
399
27016
λ(-)/nm
651
420,761,8429
42816
-0.37;-0.606
-0.12;-0.6915
-0.07;-1.0515
Figure 1. UV-vis spectra of toluene solutions of (a) T2CN4 (---)
before and (s) after addition of AuNPs and (b) AuNPs upon
progressive addition of TCNQ.
816
Analytical Chemistry, Vol. 80, No. 3, February 1, 2008

Page 3

10-20 electrons over 3600 gold atoms as calculated from
comparison with the 140-atom cluster Au140(C6)53above,10and
establishes the involvement of all the gold atoms constituting the
clusters.
TCNQ Reduction by Gold. The reaction between TCNQ and
gold does not generally occur on gold sheets. In fact, after 2 h of
dipping in a 10-3M TCNQ solution in toluene, no sign of TCNQ
radical anion formation is detected. Instead, the subsequent
addition of 0.1 M TOABr causes fast dissolution with development
of the green color (and spectral characteristics) of TCNQ radical
anion. Thus, it is suggested that TCNQ oxidizes gold with the
help of the coordinating power of bromide. Similarly, aerial oxygen
reduction by gold sheet in chloroform has been recently reported
to occur in the presence of tetralkylammonium bromide.12
To support this suggestion, we have considered the electro-
chemical dissolutive oxidation of gold metal electrode in aceto-
nitrile (Figure 2). The dissolution curve (a) is backshifted from
0.8 to 0.0 V vs Ag/Ag+by addition of 0.1 M TOABr. Although
acetonitrile (requested by electrochemistry) is clearly not quite
the same as toluene (the solvent of AuNPs), comparison with the
reversible reduction processes of TCNQ (c) suggests that gold
is in this case in the condition of reducing TCNQ to the radical
anion. At the same time, gold oxidation proceeds likely to the
first oxidized state Au(I) in the form of a dibromide anion. The
overall process may be summarized by eq 1.
CV was used for confirmation of the products of such a reaction.
Thus, AuNPs in TOABr were treated with TCNQ in a 1:1 ratio,
and after toluene evaporation, the resulting green solid was
dissolved in acetonitrile. The CV at a gold electrode (Figure 3a)
shows the one-electron reduction of the TCNQ radical anion at
E0) -0.69 V and the subsequent reduction of the gold(I)
dibromide at -1.1 V. The latter was identified by CV of anodic
products on a gold electrode (Figure 3b). Moreover, the CV of
an authentic sample of TBA+(AuBr2)-prepared according to the
literature13(Figure 3c) displays its reduction at Ep) -1.10 V,
that is, the same found in the reaction of AuNPs and TCNQ.
TCNQ Titration of AuNPs. A typical titration of AuNPs with
TCNQ (shown in Figure 5, see below) is performed as follows:
A 0.7-mL portion of a 10-4M TCNQ solution is transferred to a
0.2-cm quartz cuvette and nitrogen-degassed. TOABr-capped
AuNPs (1.2 g L-1) in toluene are used to titrate the TCNQ with
a microsyringe provided with a Teflon-coated needle. The strong
peak at 856 nm is used to monitor the formation of TCNQ radical
anion. Spectra are recorded initially and after each 1-3-µL
stepwise addition of titrant.
(12) Mortier, T.; Persoons, A.; Verbiest, T. Inorg. Chem. Commun. 2005, 8, 1075.(13) Braunstein, P.; Clark, R. J. H. J. Chem. Soc., Dalton Trans. 1973, 1845.
Figure 2. Single-sweep voltammogram for anodic dissolution of gold
sheet electrode in acetonitrile (a) + 0.1 M Bu4NClO4and (b) + 0.1 M
TOABr; (c) CV of 10-3M TCNQ in acetonitrile + 0.1 M Bu4NClO4for
comparison. Scan rate: 0.1 V s-1.
Au(0) + TCNQ + 2TOA+Br-f
TOA+(AuBr2)-+ TOA+TCNQ-(1)
Figure 3. (a) Reduction cyclic voltammogram of 3 × 10-3M AuNPs
+ 3 × 10-3M TCNQ in acetonitrile + 0.1 M TOABr on gold
minielectrode; (b) single-sweep reduction voltammogram on gold
sheet electrode in acetonitrile + 0.1 M TOABr; (c) single-sweep
reduction voltammogram of 6 × 10-3M Bu4N+AuBr2-in acetonitrile
+ 0.1 M Bu4NClO4on gold minielectrode. Scan rate: 0.1 V s-1.
Figure 4. UV-vis spectra of (s) 1.5-nm C12-Au and (- - -) 5-nm
TOABr-Au nanoparticle solutions in toluene.
Analytical Chemistry, Vol. 80, No. 3, February 1, 2008
817

Page 4

Previous degassing of the solution with nitrogen is required,
since aerial oxygen may cause some interference via back
oxidation of TCNQ radical anion during titration. Metal parts (such
as, e.g., steel needles) should be avoided, since they also may
reduce TCNQ. For the same reason, any reducing agent in the
medium must be previously excluded.
The titration has been performed at three different AuNPs
concentrations in the range 0.1-1.0 g L-1, resulting in a TCNQ/
Au ratio of 1.0 with a standard deviation of ∼5%.
So far, the procedure has been used with TOABr-capped
AuNPs in which capping is weak (by anion adsorption), and as a
consequence, all gold atoms in the nanoparticle are zerovalent.
To check the extensibility of the method to the generality of
capped AuNPs in organic media, the same procedure was used
with dodecanethiol-capped AuNP C12-Au, prepared according to
Brust7and dissolved in toluene (typically 1.43 g L-1). In this case,
capping is strong (by covalent binding of thiol to gold), and the
capped gold atoms are in the I valence state and therefore not
involved in titration. The selected nanoparticles are ∼1.5 nm in
diameter so that their SP spectrum is sluggish, showing only a
badly defined slope (Figure 4) instead of the clear peak at 525
nm of 5 nm AuNPs.
The reaction with TCNQ does not proceed unless bromide
(typically 50 mM) is added to the titrating AuNPs solution, as in
the TOABr-capped AuNPs. The strong thiol capping makes the
titration reaction slower at the beginning (e.g., ∼20 min for the
first 10% titrant addition), but if the TOABr is present in the titrated
TCNQ solution, the reaction is fast, being complete in the time
of mixing. For this reason, 0.05 M TOABr was better added to
the titrated TCNQ solutions.
The titration for the selected example, illustrated in Figure 5
(21 µL at the equivalent point), gives the dodecanethiol-capped
AuNPs (30 µg) a zerovalent gold content of 14 µg (47%), and
considering that gold(I) is present at 50% in the remaining 16 µg
of gold dodecylsulfide C12H26S-Au, the total gold content (22 µg)
is therefore 73%. The result is in good agreement with that (75%)
reported in the literature for such nanoparticles,7also considering
that we are not dealing with compounds of well-defined stoichi-
ometry but with clusters for which size and composition may be
to some extent dependent on the details of synthesis.
The titration may be performed with other anions, such as
chloride (as, e.g., the TOA salt), whereas iodide cannot be used,
since it is itself a reducing agent toward TCNQ due to its low
redox potential. Moreover, other solvents generally used for
capped AuNPs, such as chloroform, can be used with no particular
difference.
Other tetracyanoquinoidal-type molecules different from TCNQ
can, in principle, be used for such gold titration, such as TCNE
(Chart 1), which displays a reversible one-electron reduction close
to that of TCNQ (see Table 1). Yet its toxic characteristics, due
to easy hydrolysis by moist air to hydrogen cyanide, does not
recommend its use. On the contrary, TCNQ is not so hazardous
and is commonly used in chemical labs.
CONCLUSIONS
Gold nanoparticles in toluene solution reduce tetracyanoquinoi-
dal compounds such as bis(dicyanomethylene)-bithiophene and
TCNQ to the corresponding radical anions, as shown by UV-vis
spectroscopy and electrochemistry. The reaction is promoted by
bromide or chloride anions and in the case of TCNQ involves the
total amount of Au(0) in the nanoparticles. This result has allowed
the establishment of a fast and accurate spectrophotometric
method for the direct evaluation of the gold content of capped
AuNPs in organic solvents.
Although several methods are available for gold determina-
tion,14to our knowledge, there is no direct procedure for a practical
analysis of gold nanoparticles. The proposed method appears to
be generally usable, with gold nanoparticles bearing caps that
make them soluble in toluene or chloroform, such as the
commonly used alkylthiol-capped AuNPs. The method allows the
determination of the zerovalent form of gold and is particularly
useful when the nanoparticle size is lower than 2 nm so that the
surface plasmon band is not displayed any longer1and cannot
therefore be used for quantitative analysis.
ACKNOWLEDGMENT
The authors thank Dr. G. Schiavon and Dr. S. Zecchin of the
CNR for helpful discussions and S. Sitran of the CNR for his
technical assistance.
Received for review August 14, 2007. Accepted November
26, 2007.
AC701715Y
(14) Rancic, S. M.; Nikolic-Mandic, S. D.; Mandic, L. M. Anal. Chim. Acta 2005,
547, 144 and references therein.
(15) Khoo, S. B.; Foley, J. K.; Pons, S. J. Electroanal. Chem. 1986, 215, 273.
(16) Dixon, D. A.; Miller, J. S. J. Am. Chem. Soc. 1987, 109, 3656.
Figure 5. Typical spectrophotometric titration of gold nanoparticles.
10-4M TCNQ + 0.05 M TOABr in toluene (0.7 mL) is progressively
added with 1.43 g L-1C12S-Au in toluene; λ ) 856 nm; d ) 0.2 cm.
818
Analytical Chemistry, Vol. 80, No. 3, February 1, 2008