Fahlman B.D. Materials Chemistry

Подождите немного. Документ загружается.

When molecular precursors are decomposed in the gas phase, they ﬁrst exhibit

homogenous nucleation which leads to supersaturation. Condensation of sub-

nanosized particulates will ensue, followed by growth via Ostwald ripening. Particle

growth is quenched by shutting off the ﬂow of precursor, or cooling the system by

either cool-gas dilution or free-jet expansion through a narrow nozzle. The overall

size and morphology of the nanoparticles may be controlled by favoring either

nucleation or growth via sintering. That is, if sintering/annealing is faster than

collisions, large spheres will be formed; in contrast, aggregates of smaller nanopar-

ticles will be produced if nucleation processes are faster than sintering.

Figure 6.36 shows a variety of gas-phase techniques that have been used to

synthesize 0-D nanoparticles. Radio frequency plasma sources have long been

used for quantitative analysis by atomizing component species in liquid or solid

samples – a technique referred to as inductively-coupled plasma atomic emission

spectroscopy (ICP-AES). The extreme energy of an ICP may also be exploited to

vaporize precursor sources to afford the growth of nanoparticles (Figure 6.36a).

[112]

In this system, the nanoparticle size/morphology would be mostly controlled by the

concentration of precursor in the plasma, and the rate of cooling – a function of its

distance from the plasma source.

Though gas-phase techniques are not generally useful for shape-control of

nanoparticles, plasma techniques have been used to generate cubic nanocrystals.

[120]

Using a precursor gas mixture of SiH

, cubic nanocrystals are thought to be

formed due to preferential etching of spherical Si nanoparticles by H atoms formed

in the plasma, which would more readily attack (100) facets than (110) or (111) –

Figure 6.37.

As one would expect, simple ﬂame pyrolysis has been used to generate 0-D

nanoparticles. In fact, millions of metric tons of carbon black and metal oxide

nanoparticles are synthesized by this method each year. The complex ﬂame chemis-

try is difﬁcult to control, which often generates a broad distribution of nanoparticle

sizes, compositions, and morphologies. A degree of growth control may be afforded

by placing electrodes at the exit side of the ﬂame reactor. By varying the applied

ﬁeld strength, one may control the sizes and degree of agglomeration of the formed

nanoparticles.

[121]

A spark formed between two charged electrodes may also be used

to form nanoparticles (Figure 6.36b). Such spark-facilitated methods are used for

solid sources with a high melting point such as Si or C, which are not easily

evaporated in a furnace.

Laser sources are also useful for either pyrolysis of precursor vapors

(Figure 6.36c), or vaporization of solid precursor targets (Figure 6.36d ). Using a

laser (100 W), one is able to grow nanoparticles from precursor powders,

crystals, and sintered blocks of Fe

, CaTiO

,Mg

SiO

, and metal carbides.

[122]

The utility of a laser source is the introduction of localized heating, which facilitates

nanoparticle growth in controlled regions such as reactor ﬂask walls or within a

quenching cold trap. Laser ablation (Figure 6.36e) may be distinguished from laser

vaporization by the type of fragments that are produced. Whereas ablation will result

in both sub-nano and micron-sized fragments in the gas phase, vaporization results

498 6 Nanomaterials

Aperature

KBr

Window

Glass irradiation cell

Electric

feedthrough

Thermocouple

Fused

silica

window

Laser

pulse

Target

Liquid

Nitrogen

Belt

Stepping

motor

To load lock

Substrate

gas

To turbomolecular

via cold tra

laser

Fe(CO)

5(g)

nano

Air inlet

nozzle

Si electrodes

Collection

substrate

thermophorestic

collector

Figure 6.36. Illustrations of apparati used in the gas-phase synthesis of 0-D nanoparticles. Shown are:

(a) plasma system; (b) spark-facilitated growth

[113]

; (c) laser vaporization/pyrolysis

[114]

; (d) laser

evaporation

[115]

; (e) laser ablation

[116]

; (f) inert-gas evaporation

[117]

; (g) electrospray system

[118]

;

(h) spray pyrolysis.

[119]

6.3. Nanoscale Building Blocks and Applications 499

1000 mm

Quenching

Gas

Filter,

Vacuum

Pump

Cooling

Water

180 mm

Evaporation

Zone

Carrier

Gas

Gas Return Cap:

Syringe Pump

Filter

Flow meter

Dryer

75ⴗ

Compressed Air

The Capillary Tested

SMPS

EtOH

evaporation

Organic species

removed

30 nm

100 ⬚C

400 ⬚C

600 ⬚C

800 ⬚C

1000-1500 ⬚C

Hydroxy groups

removed

Grain growth and

phase change

Grain growth,

crystal habit formation

Increasing temperature

Air

81 μm I.D.

224 μm O.D.

Air

-H.V.

Droplet:

EtOH

Ti(OR)4

H2O

AcOH

Solid particle

formed via

chemical rxn

Hydrolyzed

titanium oxide

and impurities

(amorphous)

Rutile

single crystal

with possible twins

and stackin

faults

Polycrystalline

titanium dioxide of

anatase and rutile

Nanocrystalline

titanium dioxide

of anatase

Radioactive

Sources, Po

210

(10 mCi)

Flow meter

Quenching Zone

(Diluter)

100 mm 315 mm

Figure 6.36. Continued

500 6 Nanomaterials

in purely gas-phase molecular precursors. Hence, laser ablation may be considered

a type of ‘top-down’ approach, whereas laser vaporization would be a ‘bottom-up’

technique, which proceeds via homogeneous nucleation/growth.

The particle size distribution may be controlled in both laser ablation/vaporization

techniques by manipulating the pulse energy/duration (usually in the timeframe of

10–50 ns/pulse). It should be noted that when laser ablation is used for thin-ﬁlm

deposition, the technique is referred to as pulsed-laser deposition (PLD).

[123]

At lower operating pressures (ca. <0.1 Torr), thin ﬁlms will be favored; however,

at higher pressures (ca. >1 Torr), nanoparticles will be formed due to a greater

opportunity for gas-phase nucleation to occur.

Physical vapor deposition (PVD) techniques such as evaporation or sputtering

may also be used to synthesize 0-D nanoparticles. One technique, known as inert gas

evaporation (Figure 6.36f), consists of evaporation of a precursor material within a

cooling inert gas at low pressures (ca. 100 Pa). Vaporization may be accomplished

via resistive heating, ion bombardment (sputtering), or laser irradiation. As one

Figure 6.37. Schematic (top) illustrating the selective etching of facets within a Si nanoparticle by

hydrogen atoms present in a plasma. The TEM image (bottom) illustrates the resultant cubic

morphology of Si nanoparticles, realized by faster etching of Si(100) facets relative to (110) planes.

Reproduced with permission from Bapat, A.; Gatti, M.; Ding, Y. -P.; Campbell, S. A.; Kortshagen, U.

6.3. Nanoscale Building Blocks and Applications 501

might expect, reactive gases such as O

or NH

may be plumbed into the system to

allow for the production of oxide or nitride nanoparticles, respectively.

In addition to evaporating solid sources, one may also utilize dilute liquid

suspensions of nanoparticles. However, simple evaporation of these solutions in

an oven will serve to concentrate any impurities that were present in the original

solvent. In order to circumvent this limitation, the nanoparticle suspensions must be

sprayed into a heat source as small droplets whose uniformity may be controlled

using a nebulizer or electrospray techniques (Figure 6.36g). The growth of nano-

particles via this synthetic technique is referred to as spray pyrolysis, and has been

used to synthesize a variety of metal and metal oxide 0-D nanoparticles/nanoclus-

ters.

[124]

As shown in Figure 6.38h, one may easily control the crystallinity/phase of

the developing nanoparticle by varying the annealing temperature.

Solution-phase syntheses of 0-D nanostructures

Most of our discussion thus far has involved some rather extreme synthetic envi-

ronments of lasers, plasmas, or high-temperature pyrolysis. However, a pref erred

route toward nanoclusters/nanoparticles of metals and their compounds is through

use of relatively mild conditions – often taking place at room temperature on the

bench-top. This is not possibl e for carbon nanoallotropes, since the precursor (e.g.,

graphite) contains strong covalent bonding, which requires a signiﬁcant amo unt of

energy for dissociation prior to self-assembly. However, for metallic and compound

nanostructural growth, the simple reduction of metal salts (usually via NaBH

or hydrazine as reducing agents), or combination of reac tive molecular precursors,

are amenable for mild, solution-phase growth. For instance, Eqs. 6 and 7 represent

the reduction of copper ions by sodium borohydride (and subsequent hydrolysis of

borane); Eq. 8 shows the side-reaction of NaBH

with the water, indicating its

relative instability in aqueous solutions. In theory, any metal with a larger standard

reduction potential (E



) than the reducing agent (e.g ., 1.33 V for sodium borohy-

dride; 0.23 V and 1.15 V for hydrazine in acidic and basic media, respectively) is

a candidate for reduction to its metallic form. This includes most of the ﬁrst-row

transition metal ions, and many others from the main group/transition metal series.

However, it should be noted that solution pH and side-reactions (e.g., metal ions

being converted to borides by BH



rather than reduction) often provide a barrier

toward successful metal ion reduction.

CuCl

þ2NaBH

! Cu

þ 2 NaCl + B

þ H

ð6Þ

þ 6H

O ! 2 B(OHÞ

þ 6H

ð7Þ

NaBH

þ 2H

O ! NaBO

þ 4H

ð8Þ

If the above reduction were to be carried out as-is, a metallic ﬁlm or bulk powder

would be formed rather than nanostructures. That is, as the metal ions are reduced,

they would instantly agglomerate with one another to form larger particulates.

Hence, the most crucial component of nanostructure synthesis is the stabilizing

502 6 Nanomaterials

agent that isolates the growing nuclei from one another. In addition, shape control of

nanostructures may be afforded in solution by varying the stabilizing agent or other

experimental conditions.

[125]

Some desirable traits of a stabilizing (or entraining)

agent are:

(i) Chemically unreactive toward the growing nanocluster, rendering an unpassi-

vated nanocluster surface ;

(ii) Stru cturally well-deﬁned (size/shape), which allows for the controlled growth

of the encapsulated nanocluster;

(iii) Comprised of light elements (organic-based), so its structure does not interfere

with the characterization of the entrained nanocluster. This will also facilitate

its sacriﬁcial removal from the nanocluster by pyrolysis at relatively low

temperatures, if desired;

(iv) Surface-modiﬁable, to allow for tunable solubility and selective intera ctions

with external stimuli. In addi tion, to afford controllable self-assembly of

entrained nanoclusters on a variety of surfaces through chemisorption, if

desired.

Though some references cite the processing conditions (“continuous” gas-phase vs.

“batch” solution-phase) as being superior for gas-phase techniques, the resultant batch-

to-batch properties (e.g., monodispersity, morphology, purity) of solution-grown nano-

particles has greatly improved in recent years. With optimized nanoparticle synthetic

recipes now pervasive in the literature, it is now safe to assume that the nanoparticles

generated in one batch will be virtually identical to those in subsequent batches (though

one must always conﬁrm this via appropriate characterization techniques). Whereas

the nanoparticles generated from individual gas-phase runs may exhibit very similar

diameters/morphologies, their overall monodispersity will be relatively poor due to

the lack of an entraining agent (Figure 6.38). Accordingly, systems such as a differen-

tial mobility analyzers (DMAs, Figure 6.39) are often used to improve the size/

morphological uniformity of gas-phase synthesized nanoparticles.

[126]

In aqueous solutions, the most common method used to stabilize nanostructures is

the use of organic “capping” ligands. For instance, the Turkevich process, which

dates back to early colloidal growth of the 1950s (and reﬁned by Frens in the 1970s),

uses sodium citrate (I) to entrain the reduc ed gold nuclei. Particle diameters on the

order of 10–20 nm may be synthesized using this method. In this case, since gold is

easily reduced (E



¼ 1.00 V for AuCl



), the citrate reagent acts as both the reducing

and stabilizing agent.

The cationic stabilizing agent bis(11-trimethylammoniumdecanoylaminoethyl)

disulﬁde dibromide (TADDD, II) has also been utilized to grow metallic

6.3. Nanoscale Building Blocks and Applications 503

Figure 6.38. TEM images illustrating the common relative monodispersities of solvent-based (top) and

gas-phase (bottom) syntheses. The top image is of CdSe quantum dots synthesized using a solution-phase

mixture of TOPO/TOP/HDA at 300



C. Reproduced with permission from Talapin, D. V.; Rogach, A. L.;

Society. The bottom TEM image shows silicon nanoparticles generated from the laser pyrolysis of silane

gas. Reproduced with permission from Swihart, M. T. Curr. Opin. Coll. Interf. Sci. 2003, 8, 127.

504 6 Nanomaterials

nanoclusters with diameters <10 nm. If NaBH

is used as the reducing agent, the

sulﬁde bond is cleaved resulting in a –SH capping group. The thiol is chemisorbed to

the surface of the growing nanostructure surface to prevent agglomeration (esp.

effective for “thiol-philic” noble metals such as Pt, Ag, and Au).

[127]

Recently, there has been much interest in the use of structurally perfect polymeric

dendrimers (see Chapter 5) as stabilizing templates for nanocluster growth. By

varying the peripheral groups and number of repeat branching units (known as

Figure 6.39. Schematic of nanoparticle growth via pulsed-laser vaporization with controlled condensation

(LVCC), coupled to a differential mobility analyzer (DMA). A DMA is used to control the size of gas-

phase synthesized nanoparticles by exploiting differences in the electrical mobility of nanoparticles under a

ﬂow of an inert gas. Reproduced with permission from Glaspell, G.; Abdelsayed, V.; Saoud, K. M.;

gas-phase techniques may be used to synthesize Au/Ga core/shell nanoparticles with the assistance of

multiple DMAs. Reproduced with permission from Karlsson, et al. Aerosol Sci. Technol. 2004, 38, 948.

6.3. Nanoscale Building Blocks and Applications 505

“Generations”), one is able to ﬁne-tune the size of the entrained nanocluster. Though

amine-terminated dendrimers and hyperbranched polymers may be used as a tem-

plate for M

chelation and subsequent chemical reduction, the size of the resultant

nanoparticle is relatively large, with a greater degree of agglomeration possible.

This is especially the case for hyperbranched polymers that exhibit a random

structure, which results in a much greater nanoparticle polydispersity. On the

other hand, if the primary surface amines (NH

) are either protonated (NH

), or

replaced with hydroxyl groups (—OH), the prereduced metal ions are forced further

into the interior of the dendritic structure (Figure 6.40). This results in much smaller

diameters and narrow polydispersities for the reduced metal nanoclusters.

[128]

1⬚

3⬚

1⬚

3⬚ amines 1⬚ amines

3⬚

1⬚

2-SBD

2-SBD 14 16

4-SBD 62 64

6-SBD 254 256

6-SBD

pH 6

pH 8

Repeating unit:-(CH

(CO)NHCH

N)-

pKa(1⬚) ~ 7-9

(

3⬚

)

~ 3-6

Figure 6.40. Molecular structure of a second Generation (G2) amine-terminated PAMAM dendrimer,

illustrating the positions of the metal ions chelated to the primary amine groups (prereduction). In

comparison, a G2 hydroxyl-terminated PAMAM dendrimer is shown, with the metal ions now

preferring to chelate to the interior tertiary amine groups. Shown on the bottom is the effect of

protonation on G2/G6 amine-terminated PAMAM dendrimers. A schematic on the lower right illustrates

the positions of the protonated amines at varying pH values. As the generation size increases, the surface

density also increases which limits the access of protons (or chelating metal ions) to interact with the

interior tertiary amine groups. Reproduced with permission from Kleinman, M. H.; Flory, J. H.; Tomalia,

506 6 Nanomaterials

In addition to controll ing the surface moieties and solution pH, the generation

size of the dendrimer is also paramount for successful nanocluster growth. As the

degree of branching increases, s o does the s urfa ce d ens it y, whi ch pre v ents

the incoming metal ion (or H

during protonation) from entering t he interior of

the dendritic architecture. Alternatively, for smaller generations, the entrained

species becomes easily dislodged from the interior due t o its open structure.

Hence, the most effective size range for nanocluster growth using polyamidoa-

mine (PAMAM) is between the fourth and sixth generations (G4–G6), which

exhibit strong container properties (Figure 6.41). As an illust ration of the extrem e

ﬂexibility of the dendritic architecture, the core may also be altered to change its

solubility characteristics, or allow t he penetration of species through the periphery

at high generations (Figure 6.42). Since dendrimers contai ning an alm ost unlimi ted

range of cores and peripheral groups have been synthesized, it is now possible to

easily control nanocluster properties such as composition, size, morphology, solu-

bility, and encapsulation (e.g., control the release rate of entrained medicinal

agents/sensors based on structural or environmental changes, for targeted drug

delivery or in situ monitoring).

It should be noted that metal nanocluster growth using dendritic templates is

strongly governed by the degree of complexation of the precursor metal ions. For

instance, silver nanoclusters are not possible using hydroxyl-terminated PAMAM

dendrimers since Ag

is not strongly chelated to tertiary amine groups. However, if

nanoclusters are ﬁrst generated within the structure, followed by Ag

, a redox

Figure 6.41. Relative sizes and surface densities of PAMAM dendrimers, showing the most suitable

range for nanocluster growth as Generation 4 to Generation 6. Reproduced with permission from

Dendrimers and other Dendritic Polymers, Frechet, J. M. J.; Tomalia, D. A. eds., Wiley: New York, 2001.

6.3. Nanoscale Building Blocks and Applications 507