Materials Research Bulletin 55 (2014) 131–136
One-pot eco-friendly synthesis of gold nanoparticles by glycerol inalkaline medium: Role of synthesis parameters on the nanoparticlescharacteristics
Eduardo B. Ferreira, Janaina F. Gomes, Germano Tremiliosi-Filho, Luiz H.S. Gasparotto *Instituto de Química de São Carlos, Universidade de São Paulo, Caixa Postal 780, 13560-970 São Carlos, Brazil
A R T I C L E I N F O
Article history:Received 9 September 2013Received in revised form 12 March 2014Accepted 3 April 2014Available online 05 April 2014
Keywords:A. MetalsA. NanostructuresB. Chemical synthesisC. Electron microscopy
A B S T R A C T
In this work we studied the role of experimental variables in an ecologically-correct synthesis of goldnanoparticles carried out by glycerol in alkaline medium at ordinary temperatures (25 �C and 0 �C).Variation of pH allowed the production of spherical and anisotropic nanoparticles. Differentconcentrations of polyvinylpyrrolidone (PVP) directly impacted the size and stability of thenanoparticles. In contrast, glycerol concentration had little influence on the synthetic process. Theempirical rate law was determined for the process with respect to glycerol and gold ions. From the kineticstudy it was possible to establish that the rate of nanoparticle formation is only slightly more dependenton gold ions than on glycerol. Thus, to increase the rate of nanoparticles formation, it is economically andenvironmentally more advantageous to increase the glycerol concentration than the Au3+ concentration.
ã 2014 Elsevier Ltd. All rights reserved.
Contents lists available at ScienceDirect
Materials Research Bulletin
journal homepage: www.else vie r .com/ locat e/mat resbu
1. Introduction
Huge attention has been devoted to gold nanoparticles (AuNPs)due to their broad potential application in photonics [1,2],electrocatalysis [3,4], chemical sensing [5], biosensing [6] andcancer therapy [7]. Haruta [8] quoted gold as the “novel catalyst of21st century”, due to its application in CO oxidation. Popularnanogold synthesis involves reducing agents such as sodiumborohydride [9,10], sodium citrate [11] and hydrazine [12]. Inparticular, the citrate procedure is inconvenient because goldnanoparticles are formed only upon heating the solution(80–100 �C). Hydrazine is quite efficient as a reducing agent,however intrinsic drawbacks such as carcinogenicity, environ-mental hazard, and instability (specially in its anhydrous form)[13] pose a major problem to scalability. Quite recently wereported an environmentally correct route to produce gold [14]and silver [15] nanoparticles for borohydride electro-oxidation andoxygen electro-reduction, respectively, using glycerol in alkalinemedium as reducing agent. Seeking for environmentally greenroutes of nanoparticle preparation has been a trend in the past
* Corresponding author. Current address: Instituto de Química, UniversidadeFederal do Rio Grande do Norte, Lagoa Nova, 59078-970, Natal, RN, Brazil. Tel.: +5584 33422323.
E-mail addresses: [email protected] (E.B. Ferreira),[email protected] (J.F. Gomes), [email protected] (G. Tremiliosi-Filho),[email protected] (L.H. Gasparotto).
http://dx.doi.org/10.1016/j.materresbull.2014.04.0030025-5408/ã 2014 Elsevier Ltd. All rights reserved.
years [16–20]. Glycerol’s non-toxicity and biodegradability make itan excellent alternative to commonly used reducing agents.Although the generation of AuNPs by glycerol is possible, theinfluence of experimental variables (e.g. pH, stabilizer concentra-tion) on the process is still unknown.
In this paper we studied the influence of different experimentalparameters such as glycerol concentration, pH, stabilizer concen-tration and temperature on the synthetic process. It is known thatshape, size and spatial arrangement of nanomaterials stronglyimpact their properties. Therefore, it is important to determinehow the synthesis variables generate nanoparticles with distinctcharacteristics. Glycerol concentration was found to influence theleast the synthetic process. A minimal of 5.0 g/L polyvinylpyrro-lidone (PVP) is necessary to achieve complete nanoparticlestabilization. It was found that pH plays a major role in regulatingsize and shape of gold nanoparticles, being possible to obtainspherical and anisotropic nanoparticles depending on the pHemployed. Finally, we discuss the empirical rate order determinedas a function of glycerol and gold ions.
2. Experimental
2.1. Reagents and instrumentation
All chemicals (Aldrich) used in this work were of analyticalgrade and used without further purification. A Varian/Cary 5Gspectrophotometer was used to acquire UV–vis spectra of the
132 E.B. Ferreira et al. / Materials Research Bulletin 55 (2014) 131–136
AuNPs colloidal suspensions. For the transmission electronmicroscopy (TEM) experiments, copper grids coated with carbonfilm were immersed into the nanoparticle colloidal suspensionsand allowed to dry overnight in a desiccator. The grids were thenanalyzed with a TEM FEI Tecnai (200 kV accelerating voltage).
2.2. Preparation of AuNPs
In a typical experiment for AuNPs synthesis, known amounts ofpolyvinylpyrrolidone (PVP, molecular weight = 10.000) with AuCl3(30 wt% in HCl) were dissolved in 5 mL of ultrapure water. In aseparate flask, determined quantities of a glycerol and NaOH weredissolved in 5 mL of ultrapure water. The glycerol–NaOH solutionwas then added to the AuCl3–PVP solution to yield 10 mL ofsolution with final concentrations shown in Table 1. The effect ofpH was investigated at two temperatures (25 �C and 0 �C). 0 �C wasachieved with an ice bath. To follow the reaction kinetics (Table 1,Set 4#), the wavelength of 520 nm was fixed and the absorbancefollowed as a function of time with the reaction conducted directlyin the UV–vis cuvette.
3. Results and discussion
3.1. Influence of glycerol concentration
Fig. 1A shows a collection of the UV–vis spectra of the colloidalAuNPs prepared at room temperature (�25 �C) at different glycerolconcentrations with all other parameter held constant. It wasattempted to produce gold nanoparticles in absence of PVP,however, they tended to agglomerate and precipitate in the bottomof the reaction vessel. Unlike citrate and borohydride anions,glycerol itself lacks capability of stabilizing the nanoparticles.Increasing the glycerol concentration caused an increase in AuNPsconcentration, as shown by the consistent increments in absor-bance in Fig. 1A. Red-colored solutions were immediately obtainedupon mixing glycerol/NaOH with Au3+/PVP. The red color is due tothe surface plasmon band (SPB) as a result of the resonant coherentdipolar oscillations of the electron gas (electrons of the conductionband) at the surface of nanoparticles. The SPB is a valuable tool forinference about the size regime of some metal particles (e.g. Ag andAu). The colloidal AuNPs spectra had a maximum absorbance(lmax) at around 520 nm regardless of the concentration, a valuetypical for spherical gold nanoparticles [21,22]. The symmetry ofthe bands implies a fair similarity in the shape of the nanoparticlesand low degree of aggregation in the solution [23]. From theseresults one can conclude that glycerol concentration has practicallyno influence on size and distribution of AuNPs. Non-spherical goldnanoparticles display multiple SPR bands correlated with theirmultiple axes, with one of the absorption bands appearing towardsthe near infrared region [24]. The fact that the absorbance tends tofade out for wavelengths higher than 600 nm is an indicative thatthere are no other particle geometries in the solution (e.g.nanorods, pentagons, truncated triangle plates). Fig. 1B shows arepresentative TEM image of AuNPs produced at an intermediateglycerol concentration of 6.6 � 10�2mol L�1. The nanoparticles
Table 1Experimental conditions employed in this work. Temperature was kept constant at 25
Parameters
Experiments Glycerol concentration (mol L�1) PVP (g L�1)
Set 1# 6.6 � 10�4; 6.6 � 10�3; 6.6 � 10�2; 1.0 10
Set 2# 0.10 10
Set 3# 0.10 0.040; 0.2Set 4# 0.10 and 0.50 10
were spherical in shape with mean particle size of 7.6 nm � 1.4 nm,thus corroborating the UV–vis results. Fig. 1C, which was obtainedby acquiring many images as that of Fig. 1B (100 particles counted),shows that size of the AuNps lie indeed between 6 nm and 9 nm. Itcan be seen in the high-resolution TEM (inset of Fig. 1B) that theparticles are polycrystalline.
3.2. Influence of pH
Fig. 2 presents UV–vis spectra of AuNPs produced at different pHand 25 �C with the other parameters kept constant. Two groups ofspectra can be observed: one centered at about 540 nm (pH 9 and 11)and another at around 520 nm (pH 12–14). The inset of Fig. 2 displaysphotographs of the AuNPs colloidal solutions at different pH and25 �C. The higher the alkalinity, the deeper the red-wine color. For pH9 and 11 the redshift (compared to 520 nm) is a diagnostic of anincrease in the nanoparticle size [21,25]. Another contribution forthis redshift might be an incomplete reduction of metal ionsadsorbed on the surface of the nanoparticle and/or in the solution, asarguedby Moskovits et al. [22]. At relative lowalkalinity the reducingpower of glycerol must be also low, tending to increase with the pH.This is supported by the fact that at pH higher than 12 the lmax
centered back at 520 nm, which is the typical value for fully reducedgold nanoparticles. Another interesting feature is that at pH 9 and 11the absorbance does not tend to vanish at wavelengths higher than600 nm, in contrast to pH 12–14. This behavior can be attributed tothe existence of other-than-spherical shaped nanoparticles. Foranisotropic AuNPs such as pentagon, truncated triangle plate andcuboctahedron the absorption starts at around 600 nm and developpeaks in the NIR region [24]. A TEM image of AuNPs synthesized atpH 9 (bottom left) reveals significantly larger nanoparticles withtruncated triangle, hexagonal particles and nanorods, thus corrob-orating the UV–vis results. Production of anisotropic gold nano-particles can thus be achieved by simple reduction of pH. A TEMimage of AuNPs synthesized at pH 14 (bottom right) shows that athigh pH values their size can be substantially decreased, henceconfirming the qualitative UV–vis results.
The temperature was found to dramatically impact the syntheticprocess. At 0 �C (Fig. 3) the absorptions at pH 9 and 11 are practicallynon-existent compared to pH 12–14. This can also be readilyrealized in the photographs of the AuNPs colloidal solutions (Fig. 3,bottom). At pH 9 and 11 the solutions still display a pale-yellow colorcharacteristic of gold salt, denoting that the concentration of AuNPs,if any formed at all, must be very low. When zooming in on theregion between 500 nm and 700 nm for pH 9 and 11 (inset of Fig. 3)one can actually see broad reshifted bands centered at about 558 nmand 567 nm, respectively. This is an indicative that at 0 C� and pH 9and 11 the AuCl3 has been hardly reduced and the nanoparticles arelarger than those shown in Fig.1. Upon increasing the pH, the centerof the band tends to go back to 520 nm.
3.3. Impact of the PVP concentration
Fig. 4A shows UV–vis spectra acquired at distinct PVPconcentrations with all other parameters held constant. The
�C unless otherwise indicated.
pH AuCl3(mmol L�1)
13 0.509;11–14 (at 25 �C and 0 �C) 0.50
0; 1.0; 5.0 13 0.5013 0.10; 0.25; 0.50; 0.75
Fig. 1. (A) UV–vis spectra of the colloidal AuNps acquired at the following glycerol concentrations (Set 1#): 6.6 � 10�4mol L�1, 6.6 � 10�3mol L�1, 6.6 � 10�2mol L�1 and1.0 mol L�1. (B) TEM images of the colloidal AuNps. Inset: high-resolution TEM of a 9.0 nm single nanoparticle. (C) Size distribution of the AuNPs produced at the followingconditions: 6.6 � 10�2mol L�1 glycerol, 0.10 mol L�1 NaOH, 10 g L�1 PVP and 0.50 mmol L�1 AuCl3 at 25 �C.
E.B. Ferreira et al. / Materials Research Bulletin 55 (2014) 131–136 133
amount of PVP was found to influence the lmax of the absorbancecurves. At 0.040 g/L PVP the lmax is located at 539 nm, while thatfor 0.20 g/L and 1.0 g/L PVP is at 532 nm. Ultimately, the lmax wentback to 520 nm at 5.0 g/L PVP, the same maximum wavelength
Fig. 2. Collection of UV–vis spectra obtained at 25 �C and different pH (Set 2#).Inset: photographs of the AuNps colloidal solutions taken at different pH (aslabelled on the flasks) and 25 �C. TEM images of colloidal AuNps synthesized at pH 9(bottom left) and pH 14 (bottom right). Condition of synthesis: 10 g L�1 PVP,0.10 mol L�1 glycerol and 0.50 mmol L�1 AuCl3 at 25 �C.
found for 10.0 g/L PVP. As the NaOH concentration (0.10 mol L�1)was high enough to rule out pH effects, the consistent blueshiftwith increasing PVP concentration is a diagnostic of particle sizedecrease [21]. TEM images of Fig. 5 corroborate this result, in whichone can observe that the size of AuNPs becomes smaller and thedistribution narrower as the PVP concentration increases. We alsoinvestigated the stability of the colloidal solution as a function oftime. Fig. 4B shows the lmax shift as function of PVP concentrationand time (freshly prepared, 24 h, 48 h and 72 h after de synthesis).PVP concentrations were converted to logarithm to make the datagroups more discernible. With exception of 5.0 g/L PVP, for theother concentrations the lmax suffered a slight blueshift with time.Possibly, at the beginning of the process, for solutions with low PVPconcentration, the nanoparticles may be large due to the inefficientPVP capping under high reaction rate conditions. As the time goeson the rate of reaction decreases, thus smaller nanopartiles cannow be stabilized, which accounts for the slight blueshift of thelmax and the bimodal features of the particle size distribution for0.040 g/L and 0.20 g/L. Another effect that can not be ruled out isthe ripening of big particles at low PVP concentration leading to theformation of extra small nanoparticles. Dissolution of smallparticles and redeposition of gold on the surface of larger onesis possible due to the poor PVP protection. The stabilizingcapability of a 5.0 g/L PVP solution is similar to that of 10.0 g/L,therefore no lmax shift was observed over time.
3.4. Reaction kinetics
In order to gain more insight into the synthetic process westudied the reaction kinetics of the formation of nanosized goldcolloids. Global reaction orders were determined for gold andglycerol. All other parameters (pH, PVP concentration andtemperature) were kept constant at values that enabled thereaction to proceed without restriction. In this way the reactionrates are solely due to glycerol and gold ions.
Fig. 3. Top: UV–vis spectra at different pH and 0 �C (Set 2#). Inset: zoom in on theregion between 500 nm and 700 nm for pH 9 and 11. Bottom: photographs of theAuNps colloidal solutions at different pH and 0 �C. Condition of synthesis:0.10 mol L�1 glycerol, 10 g L�1 PVP and 0.50 mmol L�1 AuCl3.
Fig. 4. (A) Stable UV–vis spectra obtained at different PVP concentrations after 72 h(Set 3#). (B) Evolution of the maximum wavelength (lmax) as a function of time andPVP concentration. (!) 0.040 g/L PVP, (*) 0.20 g/L PVP, (~) 1.0 g/L PVP, (�) 5.0 g/LPVP. Condition of synthesis: 0.10 mol L�1 glycerol, 0.10 mol L�1 NaOH, 0.50 mmol L�1
AuCl3 and 25 �C.
134 E.B. Ferreira et al. / Materials Research Bulletin 55 (2014) 131–136
A general reaction between gold ions and glycerol may bewritten as follows:
xAu3þ þ yC3H8O3 ! nanoparticles þ organic products (1)
where x and y are the stoichiometric coefficients of gold ions andglycerol, respectively. The actual stoichiometric coefficients are notknown because other species, such as the deprotonated glycerol forinstance, might also play a role in the reduction. However, notknowing the stoichiometric coefficients is irrelevant as the globalrate orders may strongly depart from them.
The rate of reaction can be written as a function of any of theinvolved species (Au3+, glycerol or Au nanoparticles). We areinterested in the formation of nanoparticles, therefore the initialrate of nanoparticle formation is written as:
initial formation rate ¼ d½nanoparticles�dt
¼ k½Au3þ�n½glycerol�m
(2)
where k is the rate constant and n and m are the reaction orderswith respect to Au3+ and glycerol, respectively. We can apply themethod of initial rates in conjunction with the isolation method todetermine the reaction order with respect to gold and glycerol [26].The isolation method consists in setting the concentration of allreactants except one in large excess. In our case, there are only tworeactants, thus, just one has to be largely present. We chose to setglycerol in excess because the change of nanoparticle concentra-tion in time can be easily monitored by UV–vis at the fixedwavelength of 520 nm. The absorbance itself can be used in thetreatment because the concentration of nanoparticles is directlyproportional to the absorbance (Lambert–Beer law). As theconcentration of glycerol is much higher than those of Au3+
(glycerol concentration must be at least one hundred times higherthat those of gold ions) it changes slightly in the very beginning ofthe reaction. Thus, it can be incorporated into k generating anapparent rate constant k0 . Eq. (2) becomes:
initial formation rate ¼ d½nanoparticles�dt
¼ k0½Au3þ�n (3)
where k0 = k [glycerol]m. The left-hand of Eq. (3) is the initial ratesmeasured at the beginning of the reaction for different initialconcentrations of gold ions. Fig. 6 depicts the variation of theabsorbance with time for distinct initial Au3+ concentrations (from0.1 mmol L�1 to 0.75 mmol L�1) at fixed glycerol concentrations(0.1 mol L�1 and 0.5 M mol L�1 glycerol). It can be seen that at thebeginning of the reaction the slopes of absorbance vs. time are verysteep, which implies fast reactions within the first minute. Wedetermined the slopes (that are the initial rates, V0) of theabsorbance vs. time curves for both sets of data in Fig. 6 and plotteda log–log graph from the following equation:
log V0 ¼ logd½nanoparticles�
dt
� �¼ log k0 þ n log½Au3þ� (4)
For a series of initial concentrations of Au3+, the plot of thelogarithms of the initial rates against the logarithms of the initialconcentrations of Au3+ should be a straight line with slope n, asshown in Fig. 7. From the intercept with the y axis in Fig. 7 one canget the apparent constants for 0.5 mol L�1 and 0.1 mol L�1 glycerol,which, in turn, can be used to determine the order in relation toglycerol through k0 = k [glycerol]m. From this equation k also followsafter calculating the order with respect to glycerol. The empiricalrate law was found to be:
initial formation rate ¼ 1:9 � 10�3½Au3þ�3=4½glycerol�1=2 (5)
From Eq. (5) one can see that gold ions are only slightly moreimportant in increasing the reaction rate than glycerol. This means
Fig. 5. TEM images of AuNPs prepared at different PVP concentrations (Set 3#). Condition of synthesis: 0.10 mol L�1 glycerol, 0.10 mol L�1 NaOH, 0.50 mmol L�1 AuCl3 and25 �C.
Fig. 6. Absorbance vs. time curves for different Au3+ concentrations rangingbetween 0.10 mmol L�1 and 0.75 mmol L�1 at 0.10 mol L�1 and 0.50 mol L�1 glycerolacquired at a fixed wavelength of 520 nm (Set 4#). Condition of synthesis:0.10 mol L�1 NaOH and 10 g L�1 PVP, 25 �C.
Fig. 7. Logarithm of the initial nanoparticle formation rate vs. the logarithm of Au3+
concentration (Set 4#). These curves were used for extraction of reaction orders andthe rate constant. Condition of synthesis: 0.10 mol L�1 NaOH and 10 g L�1 PVP, 25 �C.
E.B. Ferreira et al. / Materials Research Bulletin 55 (2014) 131–136 135
that, to achieve higher reaction rates, it is more interesting toincrease the glycerol concentration than gold concentration as thelatter is far more expensive than the former.
4. Conclusions
The eco-friendly synthesis of gold nanoparticles by glycerol inalkaline medium at room temperature is dependent on the
experimental parameters. The glycerol concentration did not affectthe position of the maximum wavelength in the UV–vis experi-ments. By varying the pH one can obtain either spherical oranisotropic AuNPs. Low temperature was found to be detrimentalfor the process. The stabilizer PVP also played a role in thecharacteristics of the AuNPs. From the kinetic study it wasdetermined that the rate of nanoparticle formation is slightly moredependent on gold ions than on glycerol. The straightforwardnessof this method might enable it to be routinely employed forproducing Au nanoparticles in standard laboratories. In addition itis simple, inexpensive and environmentally friendly.
Acknowledgements
The authors thank CNPq and FAPESP for the overall support ofthis research.
References
[1] D. Pissuwan, S. Valenzuela, M. Killingsworth, X. Xu, M. Cortie, Journal ofNanoparticle Research 9 (2007) 1109–1124.
[2] E. Hutter, J.H. Fendler, Advanced Materials 16 (2004) 1685–1706.[3] S.J. Guo, E.K. Wang, Analytica Chimica Acta 598 (2007) 181–192.[4] M.S. El-Deab, T. Ohsaka, Electrochemistry Communications 4 (2002) 288–292.
136 E.B. Ferreira et al. / Materials Research Bulletin 55 (2014) 131–136
[5] C.J. Murphy, A.M. Gole, S.E. Hunyadi, J.W. Stone, P.N. Sisco, A. Alkilany, B.E.Kinard, P. Hankins, Chemical Communications (2008) 544–557.
[6] G. Jerkiewicz, M. DeBlois, Z. Radovic-Hrapovic, J.P. Tessier, F. Perreault, J.Lessard, Langmuir 21 (2005) 3511–3520.
[7] I.H. El-Sayed, X. Huang, M.A. El-Sayed, Nano Letters 5 (2005) 829–834.[8] M. Haruta, Gold Bulletin 37 (2004) 27–36.[9] Y.G. Sun, B. Mayers, Y.N. Xia, Nano Letters 3 (2003) 675–679.
[10] M. Brust, M. Walker, D. Bethell, D.J. Schiffrin, R. Whyman, Journal of theChemical Society, Chemical Communications 7 (1994) 801–802.
[11] X.H. Ji, X.N. Song, J. Li, Y.B. Bai, W.S. Yang, X.G. Peng, Journal of the AmericanChemical Society 129 (2007) 13939–13948.
[12] D.H. Chen, S.H. Wu, Chemistry of Materials 12 (2000) 1354–1360.[13] U.B. Demirci, Journal of Power Sources 169 (2007) 239–246.[14] L.H.S. Gasparotto, A.C. Garcia, J.F. Gomes, G. Tremiliosi-Filho, Journal of Power
Sources 218 (2012) 73–78.[15] A.C. Garcia, L.H.S. Gasparotto, J.F. Gomes, G. Tremiliosi-Filho, Electrocatalysis 3
(2012) 147–152.[16] M. Anderson, E. Torres-Chavolla, B. Castro, E. Alocilja, Journal of Nanoparticle
Research 13 (2011) 2843–2851.
[17] S. Iravani, Green Chemistry 13 (2011) 2638–2650.[18] P. Raveendran, J. Fu, S.L. Wallen, Journal of the American Chemical Society 125
(2003) 13940–13941.[19] A.N. Grace, K. Pandian, Colloids and Surfaces A: Physicochemical and
Engineering Aspects 290 (2006) 138–142.[20] A.N. Grace, K. Pandian, Materials Chemistry and Physics 104 (2007) 191–198.[21] M.C. Daniel, D. Astruc, Chemical Reviews 104 (2004) 293–346.[22] M. Moskovits, I. Srnova-Sloufova, B. Vlckova, Journal of Chemical Physics 116
(2002) 10435–10446.[23] N. Vigneshwaran, R.P. Nachane, R.H. Balasubramanya, P.V. Varadarajan,
Carbohydrate Research 341 (2006) 2012–2018.[24] G. Zhang, J. Jasinski, J. Howell, D. Patel, D. Stephens, A. Gobin, Nanoscale
Research Letters 7 (2012) 337.[25] S. Link, M.A. El-Sayed, The Journal of Physical Chemistry B 103 (1999) 4212–
4217.[26] P. Atkins, J.D. Paula, Physical Chemistry, seventh ed., Oxford University Press,
Oxford, 2009.
Recommended
