**3. Elaboration of nanoporous copper via the chemical composition design**

unit cells. The small peak around 36° was considered to be from the Cu<sup>2</sup>

the Pd or Pt phase in the XRD patterns of dealloyed Ti60Cu39Pd<sup>1</sup>

**Figure 2.** Dependence of the grain sizes (a) and the lattice constants (b) of dealloyed Ti60Cu39M<sup>1</sup>

in G-I and G-II. Reproduced with permission from Dan et al. [30] Copyright Materials Transactions.

and Au and G-II Ni, Pd, and Pt) ribbons on the addition of the elements. The lines are used for view guide of data points

Ti60Cu39Au<sup>1</sup>

and Ti60Cu39Au<sup>1</sup>

of strong diffraction peaks from crystalline Pd and Pt phases indicates that CuPd or CuPt solid solution formed after dealloying. These XRD patterns were similar to those of Ag-, Au-, and Ni-stabilized NPCs [26–29]. **Figure 2** shows the change in the lattice constant, *α*, and the grain sizes, *L*, with 1 at% addition of Au-group (G-I) and Pt-group (G-II). The standard lattice constants are reported to be 0.3524 nm for Ni, 0.3608 nm for Cu, 0.3891 nm for Pd, 0.3924 nm for Pt, 0.4078 nm for Au, and 0.4085 nm for Ag, respectively. The lattice constants of NPCs were estimated to be 0.3615–0.3627 nm on the basis of the XRD data of Cu (111) peaks. The lattice constants became larger when the added elements had larger lattice constants except in the case of Ag. As indicated in **Figure 2**, the Cu lattice expanded more when Pd and Pt were micro-alloyed into the Ti60Cu40 alloy. The added Ni, Pd, or Pt atoms were thus considered to substitute the Cu atoms in the Cu lattice to a higher extent than in the Ag- or Au-added cases, resulting in the expansion of the Cu lattice constants from 0.3615 nm to 0.3627 nm in **Figure 2**. In previous works it was found that the Au or Ag phase was formed after dealloying the

Elaboration of Nanoporous Copper via Chemical Composition Design of Amorphous Precursor Alloys

the large expansion of the lattice constants of the NPCs indicated that more Pt-group atoms invaded the Cu lattice than Au-group atoms (i.e., Ag addition and Au addition) for NPCs from precursor alloys micro-alloyed with LSD elements [26–29]. On the other hand, the grain sizes were reduced to 15 nm and 13 nm when the Pd and Pt were added into Ti60Cu40 ribbons.

precursor alloys [26–29]. The absence of diffraction peaks from

and Ti60Cu39Pt<sup>1</sup>

O phase. The absence

43

http://dx.doi.org/10.5772/intechopen.77222

ribbons and

(M: LSD—G-I Cu, Ag,

### **3.1. Characteristics of LSD-added Ti-Cu precursor alloys and nanoporous counterparts**

The XRD patterns of the as-spun Ti60Cu40, Ti60Cu39Ni<sup>1</sup> , Ti60Cu39Pd<sup>1</sup> , and Ti60Cu39Pt<sup>1</sup> ribbons present one strong and broad diffraction peak at 41° and a weak diffraction peak at 70–75°, indicating an amorphous structure, as shown in **Figure 1**. Other LSD-substituted Ti60Cu40 precursor alloys were still amorphous states, indicating that the micro-alloying of 1 at.% LSD elements had not generated the heterogeneous microstructure or phase segregation. The high-resolution TEM images (HRTEM) and corresponding selective area diffraction patterns (SADP) of as-spun Ti60Cu40 and Ti60Cu39M<sup>1</sup> show that as-spun LSD-substituted Ti60Cu39M<sup>1</sup> alloys had an amorphous structure without the crystalline clusters and phase segregations [27]. As is well known, because G-I LSD elements, Ag and Au, have a similar crystal structure to Cu, the addition of minor amounts of G-I elements has thus no significant influence on the structure of as-spun alloys [26, 27]. In G-II elements, Ni, Pd, and Pt, three strong diffraction peaks were observed in the XRD patterns of the dealloyed Ti60Cu39Pd<sup>1</sup> and Ti60Cu39Pt<sup>1</sup> ribbons after dealloying for 43.2 ks in the 0.03 M HF solution. The diffraction peaks were identified to be *fcc* Cu and had slightly shifted to low diffraction angles due to the expansion of the LSD-substituted

**Figure 1.** XRD patterns of as-spun (a, b, c, d) and dealloyed (e, f, g, h) Ti60Cu40 (a, e), Ti60Cu39Ni<sup>1</sup> (b, f), Ti60Cu39Pd<sup>1</sup> (c, g), and Ti60Cu39Pt<sup>1</sup> (d, h) ribbons after immersion in 0.03 M HF solution for 43.2 ks. Reproduced with permission from Dan et al. [30] Copyright Materials Transactions.

unit cells. The small peak around 36° was considered to be from the Cu<sup>2</sup> O phase. The absence of strong diffraction peaks from crystalline Pd and Pt phases indicates that CuPd or CuPt solid solution formed after dealloying. These XRD patterns were similar to those of Ag-, Au-, and Ni-stabilized NPCs [26–29]. **Figure 2** shows the change in the lattice constant, *α*, and the grain sizes, *L*, with 1 at% addition of Au-group (G-I) and Pt-group (G-II). The standard lattice constants are reported to be 0.3524 nm for Ni, 0.3608 nm for Cu, 0.3891 nm for Pd, 0.3924 nm for Pt, 0.4078 nm for Au, and 0.4085 nm for Ag, respectively. The lattice constants of NPCs were estimated to be 0.3615–0.3627 nm on the basis of the XRD data of Cu (111) peaks. The lattice constants became larger when the added elements had larger lattice constants except in the case of Ag. As indicated in **Figure 2**, the Cu lattice expanded more when Pd and Pt were micro-alloyed into the Ti60Cu40 alloy. The added Ni, Pd, or Pt atoms were thus considered to substitute the Cu atoms in the Cu lattice to a higher extent than in the Ag- or Au-added cases, resulting in the expansion of the Cu lattice constants from 0.3615 nm to 0.3627 nm in **Figure 2**. In previous works it was found that the Au or Ag phase was formed after dealloying the Ti60Cu39Au<sup>1</sup> and Ti60Cu39Au<sup>1</sup> precursor alloys [26–29]. The absence of diffraction peaks from the Pd or Pt phase in the XRD patterns of dealloyed Ti60Cu39Pd<sup>1</sup> and Ti60Cu39Pt<sup>1</sup> ribbons and the large expansion of the lattice constants of the NPCs indicated that more Pt-group atoms invaded the Cu lattice than Au-group atoms (i.e., Ag addition and Au addition) for NPCs from precursor alloys micro-alloyed with LSD elements [26–29]. On the other hand, the grain sizes were reduced to 15 nm and 13 nm when the Pd and Pt were added into Ti60Cu40 ribbons.

**3. Elaboration of nanoporous copper via the chemical composition** 

ent one strong and broad diffraction peak at 41° and a weak diffraction peak at 70–75°, indicating an amorphous structure, as shown in **Figure 1**. Other LSD-substituted Ti60Cu40 precursor alloys were still amorphous states, indicating that the micro-alloying of 1 at.% LSD elements had not generated the heterogeneous microstructure or phase segregation. The high-resolution TEM images (HRTEM) and corresponding selective area diffraction patterns (SADP) of

amorphous structure without the crystalline clusters and phase segregations [27]. As is well known, because G-I LSD elements, Ag and Au, have a similar crystal structure to Cu, the addition of minor amounts of G-I elements has thus no significant influence on the structure of as-spun alloys [26, 27]. In G-II elements, Ni, Pd, and Pt, three strong diffraction peaks were

loying for 43.2 ks in the 0.03 M HF solution. The diffraction peaks were identified to be *fcc* Cu and had slightly shifted to low diffraction angles due to the expansion of the LSD-substituted

, Ti60Cu39Pd<sup>1</sup>

show that as-spun LSD-substituted Ti60Cu39M<sup>1</sup>

, and Ti60Cu39Pt<sup>1</sup>

and Ti60Cu39Pt<sup>1</sup>

ribbons pres-

alloys had an

ribbons after deal-

(b, f), Ti60Cu39Pd<sup>1</sup>

(c, g),

**3.1. Characteristics of LSD-added Ti-Cu precursor alloys and nanoporous** 

The XRD patterns of the as-spun Ti60Cu40, Ti60Cu39Ni<sup>1</sup>

observed in the XRD patterns of the dealloyed Ti60Cu39Pd<sup>1</sup>

**Figure 1.** XRD patterns of as-spun (a, b, c, d) and dealloyed (e, f, g, h) Ti60Cu40 (a, e), Ti60Cu39Ni<sup>1</sup>

(d, h) ribbons after immersion in 0.03 M HF solution for 43.2 ks. Reproduced with permission from Dan

as-spun Ti60Cu40 and Ti60Cu39M<sup>1</sup>

**design**

**counterparts**

42 New Uses of Micro and Nanomaterials

and Ti60Cu39Pt<sup>1</sup>

et al. [30] Copyright Materials Transactions.

**Figure 2.** Dependence of the grain sizes (a) and the lattice constants (b) of dealloyed Ti60Cu39M<sup>1</sup> (M: LSD—G-I Cu, Ag, and Au and G-II Ni, Pd, and Pt) ribbons on the addition of the elements. The lines are used for view guide of data points in G-I and G-II. Reproduced with permission from Dan et al. [30] Copyright Materials Transactions.

The decrease in the grain sizes was considered to be due to the retardation of the self-diffusion of Cu adatoms by LSD elements [27, 29, 30]. The diffusion distance of Cu adatoms under free diffusion patterns is prevailed in a long-distance diffusion mode [2, 5, 10, 19–21]. However, the long-distance self-diffusion of Cu adatoms was interrupted and restricted by the LSD adatoms during the rearrangement of adatoms and resulted in an accumulation of Cu and LSD adatoms in a smaller scale. Consequently, smaller grains were formed when the Pt-group elements (Pd, Pt) and Au element were used to stabilize NPCs.

The nanoporous surface morphologies of dealloyed Ti60Cu39M<sup>1</sup> (M: Cu, Ag, Au, Ni, Pd, Pt) alloys are shown in **Figure 3**. The mean pore size of the nanoporous structure of as-dealloyed Ti60Cu40 ribbons was about 71 nm after dealloying in 0.03 M HF solution. The characteristic length scales of ligaments of NPC obtained from Ti60Cu40 alloy were 74 nm in 0.03 M HF solution. The characteristic sizes for the nanopores and ligaments were confirmed to be for 41 and 48 nm for dealloyed Ti60Cu39Ag<sup>1</sup> precursor, 16 nm and 27 nm for Ti60Cu39Au<sup>1</sup> precursor, 12 nm and 26 nm for Ti60Cu39Ni<sup>1</sup> precursor, 9 nm and 24 nm for Ti60Cu39Pd<sup>1</sup> precursor, and 8.5 nm and 31 nm for Ti60Cu39Pt<sup>1</sup> precursor. **Figure 4** shows a typical bright field TEM image (BFI), their corresponding selected area diffraction pattern (SADP) of Ti60Cu39M<sup>1</sup> (M: Cu, Ag, Au, Ni, Pd, Pt) amorphous alloys after dealloying for 43.2 ks. A bi-continuous porous microstructure was formed with a characteristic pore size of 7 nm for the dealloyed Ti60Cu39Pd<sup>1</sup> ribbon and ca. 6 nm for the dealloyed Ti60Cu39Pt<sup>1</sup> ribbon, respectively (**Figure 4e** and **3f**). The diffraction rings in the SAD patterns were assigned to Cu (111), (200), (220), and (311) (JCPDS

card No. 02-1225), and the inner ring was done for Cu<sup>2</sup>

**Figure 4.** Typical bright field TEM images of dealloyed Ti60Cu40 (a), Ti60Cu39Ag<sup>1</sup>

the nanopores and ligaments of dealloyed Ti60Cu39Pd<sup>1</sup>

The ligament sizes were confirmed to be ca. 23 nm for the dealloyed Ti60Cu39Pd<sup>1</sup>

nanoporous structures with a finer porosity in comparison to those in the surface regions. It is considered that the finer nanoporous structure in the internal parts mainly resulted from the lower concentration gradients inside the channels. As shown in **Figure 4e** and **f**, the pore size of Pd-stabilized and Pt-stabilized NPCs had a mean pore size of 7 and 6 nm, respectively. As shown in **Figure 4b**–**d**, the characteristic pore size of NPCs stabilized by the micro-alloying of Ni, Ag, and Au [27–29] has been reported to be 11 nm, 28 nm, and 12 nm on the basis of TEM analysis, respectively. The mean pore size of dealloyed Ti60Cu40 ribbons confirmed by TEM observation was 39 nm in **Figure 4a** [20]. The pore size decreased more than one order due to the addition of either Pd or Pt as shown in **Figure 4e** and **f**. The characteristic scale length of

corresponding selected area diffraction patterns. Reproduced with permission from Dan et al. [30] Copyright Materials

matically. On the basis of XRD, TEM, and SEM-EDX analysis, the residue is considered to be *fcc* CuPd and *fcc* CuPt solid solution. The similarity existed in all LSD-substituted Ti60Cu39M<sup>1</sup> (M: Ag, Au, Ni, Pd, Pt). Commonly the Cu(LSD) solid solutions can be regarded as the residual phases after dealloying. However, the selective dissolution is slightly different which caused

The diffraction rings from the Cu<sup>2</sup>

(e), and Ti60Cu39Pt<sup>1</sup>

(d),Ti60Cu39Pd<sup>1</sup>

Transactions.

ca. 30 nm for the dealloyed Ti60Cu39Pt<sup>1</sup>

O (111) (JCPDS card No. 74-1230).

(b), Ti60Cu39Au<sup>1</sup>

http://dx.doi.org/10.5772/intechopen.77222

45

ribbon and

(c), Ti60Cu39Ni<sup>1</sup>

ribbons decreased dra-

O phase were absent in the inset SADP in **Figure 4e**.

(f) precursors after dealloying in 0.03 M HF solution for 43.2 ks. The insets are their

Elaboration of Nanoporous Copper via Chemical Composition Design of Amorphous Precursor Alloys

and Ti60Cu39Pt<sup>1</sup>

ribbon. The BFI images mainly reflected the internal

**Figure 3.** Typical surface morphologies of dealloyed Ti60Cu40 (a), Ti60Cu39Ag<sup>1</sup> (b), Ti60Cu39Au<sup>1</sup> (c), Ti60Cu39Ni<sup>1</sup> (d),Ti60Cu39Pd<sup>1</sup> (e), and Ti60Cu39Pt<sup>1</sup> (f) precursors after dealloying in 0.03 M HF solution for 43.2 ks. Reproduced with permission from Dan et al. [30] Copyright Materials Transactions. Reproduced with permission from Dan et al. [27, 30] copyright Elsevier and Materials Transactions.

Elaboration of Nanoporous Copper via Chemical Composition Design of Amorphous Precursor Alloys http://dx.doi.org/10.5772/intechopen.77222 45

The decrease in the grain sizes was considered to be due to the retardation of the self-diffusion of Cu adatoms by LSD elements [27, 29, 30]. The diffusion distance of Cu adatoms under free diffusion patterns is prevailed in a long-distance diffusion mode [2, 5, 10, 19–21]. However, the long-distance self-diffusion of Cu adatoms was interrupted and restricted by the LSD adatoms during the rearrangement of adatoms and resulted in an accumulation of Cu and LSD adatoms in a smaller scale. Consequently, smaller grains were formed when the Pt-group

alloys are shown in **Figure 3**. The mean pore size of the nanoporous structure of as-dealloyed Ti60Cu40 ribbons was about 71 nm after dealloying in 0.03 M HF solution. The characteristic length scales of ligaments of NPC obtained from Ti60Cu40 alloy were 74 nm in 0.03 M HF solution. The characteristic sizes for the nanopores and ligaments were confirmed to be for 41

Au, Ni, Pd, Pt) amorphous alloys after dealloying for 43.2 ks. A bi-continuous porous microstructure was formed with a characteristic pore size of 7 nm for the dealloyed Ti60Cu39Pd<sup>1</sup>

diffraction rings in the SAD patterns were assigned to Cu (111), (200), (220), and (311) (JCPDS

(BFI), their corresponding selected area diffraction pattern (SADP) of Ti60Cu39M<sup>1</sup>

precursor, 16 nm and 27 nm for Ti60Cu39Au<sup>1</sup>

precursor. **Figure 4** shows a typical bright field TEM image

(b), Ti60Cu39Au<sup>1</sup>

(f) precursors after dealloying in 0.03 M HF solution for 43.2 ks. Reproduced with permission from

Dan et al. [30] Copyright Materials Transactions. Reproduced with permission from Dan et al. [27, 30] copyright Elsevier

(c), Ti60Cu39Ni<sup>1</sup>

(d),Ti60Cu39Pd<sup>1</sup>

ribbon, respectively (**Figure 4e** and **3f**). The

precursor, 9 nm and 24 nm for Ti60Cu39Pd<sup>1</sup>

(M: Cu, Ag, Au, Ni, Pd, Pt)

precursor,

(M: Cu, Ag,

precursor, and

elements (Pd, Pt) and Au element were used to stabilize NPCs.

The nanoporous surface morphologies of dealloyed Ti60Cu39M<sup>1</sup>

and 48 nm for dealloyed Ti60Cu39Ag<sup>1</sup>

ribbon and ca. 6 nm for the dealloyed Ti60Cu39Pt<sup>1</sup>

**Figure 3.** Typical surface morphologies of dealloyed Ti60Cu40 (a), Ti60Cu39Ag<sup>1</sup>

(e), and Ti60Cu39Pt<sup>1</sup>

and Materials Transactions.

12 nm and 26 nm for Ti60Cu39Ni<sup>1</sup>

44 New Uses of Micro and Nanomaterials

8.5 nm and 31 nm for Ti60Cu39Pt<sup>1</sup>

**Figure 4.** Typical bright field TEM images of dealloyed Ti60Cu40 (a), Ti60Cu39Ag<sup>1</sup> (b), Ti60Cu39Au<sup>1</sup> (c), Ti60Cu39Ni<sup>1</sup> (d),Ti60Cu39Pd<sup>1</sup> (e), and Ti60Cu39Pt<sup>1</sup> (f) precursors after dealloying in 0.03 M HF solution for 43.2 ks. The insets are their corresponding selected area diffraction patterns. Reproduced with permission from Dan et al. [30] Copyright Materials Transactions.

card No. 02-1225), and the inner ring was done for Cu<sup>2</sup> O (111) (JCPDS card No. 74-1230). The diffraction rings from the Cu<sup>2</sup> O phase were absent in the inset SADP in **Figure 4e**. The ligament sizes were confirmed to be ca. 23 nm for the dealloyed Ti60Cu39Pd<sup>1</sup> ribbon and ca. 30 nm for the dealloyed Ti60Cu39Pt<sup>1</sup> ribbon. The BFI images mainly reflected the internal nanoporous structures with a finer porosity in comparison to those in the surface regions. It is considered that the finer nanoporous structure in the internal parts mainly resulted from the lower concentration gradients inside the channels. As shown in **Figure 4e** and **f**, the pore size of Pd-stabilized and Pt-stabilized NPCs had a mean pore size of 7 and 6 nm, respectively. As shown in **Figure 4b**–**d**, the characteristic pore size of NPCs stabilized by the micro-alloying of Ni, Ag, and Au [27–29] has been reported to be 11 nm, 28 nm, and 12 nm on the basis of TEM analysis, respectively. The mean pore size of dealloyed Ti60Cu40 ribbons confirmed by TEM observation was 39 nm in **Figure 4a** [20]. The pore size decreased more than one order due to the addition of either Pd or Pt as shown in **Figure 4e** and **f**. The characteristic scale length of the nanopores and ligaments of dealloyed Ti60Cu39Pd<sup>1</sup> and Ti60Cu39Pt<sup>1</sup> ribbons decreased dramatically. On the basis of XRD, TEM, and SEM-EDX analysis, the residue is considered to be *fcc* CuPd and *fcc* CuPt solid solution. The similarity existed in all LSD-substituted Ti60Cu39M<sup>1</sup> (M: Ag, Au, Ni, Pd, Pt). Commonly the Cu(LSD) solid solutions can be regarded as the residual phases after dealloying. However, the selective dissolution is slightly different which caused the formation of Ag and Au phases [27, 28]. Although NPCs from Ti60Cu39Ag<sup>1</sup> precursor had a large final nanopore, the elaborating behavior still happened at the initial dealloying stages with a high refining efficiency [28]. As shown in **Figure 5**, the distribution of Au LSD element was profiled. The uniform distribution of Au elements can be confirmed here. For other precursor alloys, the residual LSD elements had similar profiles which benefited from the uniform distribution of these LSDs in the amorphous precursor alloys.

#### **3.2. Effects of LSDs on surface diffusion**

On the basis of the surface diffusion-controlled coarsening mechanism, the surface diffusivity, *D*<sup>s</sup> , at various dealloying temperatures was estimated by the equation [42]:

$$D\_s = \frac{[d(t)]^\ast kT}{32\gamma t \, a^\ast} \tag{1}$$

can be modified. As shown in **Figure 5b**, the value of *D*<sup>s</sup>

with permission from Dan et al. [30] Copyright Materials Transactions.

magnitude due to the addition of Pd and Pt. Compared with *D*<sup>s</sup>

2.0 × 10−21 m2 s−1 for Ti60Cu39Ni<sup>1</sup>

*<sup>R</sup>* <sup>=</sup> *Ds*

NPCs, the decrease in *D*<sup>s</sup>

for Ti60Cu39Pt<sup>1</sup>

(denominator):

for Ti60Cu39Pt<sup>1</sup>

Ti60Cu39Au<sup>1</sup>

loyed Ti60Cu39Ag<sup>1</sup>

Ti60Cu40 ribbon, 6.0 × 10−20 m2 s−1 for Ti60Cu40 ribbon, 9.7 × 10−21 m2 s−1 for Ti60Cu39Au<sup>1</sup>

**Figure 6.** Mean pore size (a), surface diffusivity (b), and refining factor of dealloyed Ti60Cu39M<sup>1</sup>

The refining factor, *R*, is defined as the ratio between the surface diffusivity, *D*<sup>s</sup>

*Cu* \_\_\_ *Ds*

As shown in **Figure 6c**, the estimated refining factor was 968 for Ti60Cu39Pd<sup>1</sup>

ribbon, 111 for the dealloyed Ti60Cu39Ni<sup>1</sup>

ribbon, 2.3 × 10−22 m2 s−1 for Ti60Cu39Pd<sup>1</sup>

elements including Ag, Au, Ni, Pd, and Pt. The lines are used for view guide of data points in G-I and G-II. Reproduced

in the Pt-group-stabilized NPCs was remarkable.

obtained from amorphous Ti60Cu40 precursor (numerator) and NPCs stabilized by LSDs

*<sup>L</sup>* <sup>≈</sup> [ *<sup>d</sup>* (*<sup>t</sup>* )*Cu* ]4 \_\_\_\_\_\_\_ [ *<sup>d</sup>* (*<sup>t</sup>* )*<sup>L</sup>* ]4 <sup>×</sup> [ *aCu* ]4

elements into the precursor Ti60Cu40 alloy was one order higher than that of the Au-group cases. The lower value of *R* for Ti60Cu40 alloy stabilized by Au-group (G-I) elements could take into account from several aspects: (1) the larger self-diffusion coefficients of Ag and Au than that of Cu [44–48] and (2) the large misfit in the atomic radii [49]. As has been reported before [44–48], the self-diffusion coefficients of Au- and Pt-group elements are ranked as following: Ag > Cu > Au > Ni > Pd > Pt. As such, there are more possibilities for Ag adatoms to meet with other Ag adatoms during the diffusion and rearrangements processes of dealloying to form clusters and to develop the Ag phase because it has a larger self-diffusion coefficient than Cu atoms [44, 48]. The self-diffusion coefficient of Au is slightly smaller than that of Cu [45]. Because the misfit in the atomic radius between Cu and Ag and Cu and Au is about 13%, the incorporation of Ag-Au atoms into the Cu lattice becomes more difficult than that of Cu/Cu atoms. On the other hand, the self-diffusion coefficients of Pt-group (G-II)

ribbon, respectively. The surface diffusivity decreased more than four orders of

Elaboration of Nanoporous Copper via Chemical Composition Design of Amorphous Precursor Alloys

\_\_\_\_\_

ribbon, respectively. In other cases, the value of *R* changed from 4 for the deal-

ribbon, respectively. The refining efficiency of the micro-alloying of the Pt-group

was estimated to be 2.5 × 10−18 m2 s−1 for

http://dx.doi.org/10.5772/intechopen.77222

[ *<sup>a</sup> <sup>L</sup>* ]4 (2)

ribbon, and 231 for the dealloyed

ribbon, and 1.3 × 10−22 m2 s−1

of the Au-group-stabilized

ribbon and

47

. M is the selected LSD

, of NPC

ribbon and 1780

where *k* is Boltzmann constant (1.3806 × 10−23 J K−1), *γ* is surface energy, *t* is the dealloying time (43,200 s), *d*(*t*) is the pore size at *t*, *T* is the temperature, and *α* is the lattice constant. The pore size of NPCs confirmed by TEM micrographs and lattice constants calculated from XRD data were adopted for calculation of *D*<sup>s</sup> . The surface energy of Cu has been reported to be 1.79 J m−2 [40, 43]. The surface energy of micro-alloyed elements has been reported to be 1.24 J m−2 for Ag, 1.50 J m−2 for Au, 2.0 J m−2 for Ni, 2.0 J m−2 for Pd, and 2.49 J m−2 for Pt [40, 43]. The concentration of Au-group elements (Ag, Au) and Pt-group elements (Ni, Pd, Pt) in the precursor alloys was 1 at%. The concentration of added elements (Ag, Au, Ni, Pd, Pt) in NPCs after dealloying should theoretically be 2.5 at% if the dissolution of Cu in HF solution is not considered. Therefore, the surface energy of LSD-stabilized NPCs is considered to be very close to that of Cu. The surface energy of Cu, 1.79 J m−2, was adopted for the calculation of *D*<sup>s</sup> . The characteristic nanopore sizes were summarized in **Figure 6a**, which has been illustrated above. On the other hand, the surface diffusivities of these LSD-stabilized alloys in 0.03 M HF solutions

**Figure 5.** High-angle annular dark field scanning TEM image (a) and elemental mapping of Cu (b) and Au (c) of dealloyed Ti60Cu39Au<sup>1</sup> precursors after dealloying in 0.03 M HF solution for 43.2 ks. Reproduced with permission from Dan et al. [27] Copyright Elsevier.

Elaboration of Nanoporous Copper via Chemical Composition Design of Amorphous Precursor Alloys http://dx.doi.org/10.5772/intechopen.77222 47

the formation of Ag and Au phases [27, 28]. Although NPCs from Ti60Cu39Ag<sup>1</sup> precursor had a large final nanopore, the elaborating behavior still happened at the initial dealloying stages with a high refining efficiency [28]. As shown in **Figure 5**, the distribution of Au LSD element was profiled. The uniform distribution of Au elements can be confirmed here. For other precursor alloys, the residual LSD elements had similar profiles which benefited from the

On the basis of the surface diffusion-controlled coarsening mechanism, the surface diffusiv-

where *k* is Boltzmann constant (1.3806 × 10−23 J K−1), *γ* is surface energy, *t* is the dealloying time (43,200 s), *d*(*t*) is the pore size at *t*, *T* is the temperature, and *α* is the lattice constant. The pore size of NPCs confirmed by TEM micrographs and lattice constants calculated from

be 1.79 J m−2 [40, 43]. The surface energy of micro-alloyed elements has been reported to be 1.24 J m−2 for Ag, 1.50 J m−2 for Au, 2.0 J m−2 for Ni, 2.0 J m−2 for Pd, and 2.49 J m−2 for Pt [40, 43]. The concentration of Au-group elements (Ag, Au) and Pt-group elements (Ni, Pd, Pt) in the precursor alloys was 1 at%. The concentration of added elements (Ag, Au, Ni, Pd, Pt) in NPCs after dealloying should theoretically be 2.5 at% if the dissolution of Cu in HF solution is not considered. Therefore, the surface energy of LSD-stabilized NPCs is considered to be very close to that of Cu. The surface energy of Cu, 1.79 J m−2, was adopted for the calculation of *D*<sup>s</sup>

characteristic nanopore sizes were summarized in **Figure 6a**, which has been illustrated above. On the other hand, the surface diffusivities of these LSD-stabilized alloys in 0.03 M HF solutions

**Figure 5.** High-angle annular dark field scanning TEM image (a) and elemental mapping of Cu (b) and Au (c) of

precursors after dealloying in 0.03 M HF solution for 43.2 ks. Reproduced with permission from

<sup>32</sup>*γ<sup>t</sup> <sup>a</sup>* <sup>4</sup> (1)

. The surface energy of Cu has been reported to

. The

, at various dealloying temperatures was estimated by the equation [42]:

uniform distribution of these LSDs in the amorphous precursor alloys.

**3.2. Effects of LSDs on surface diffusion**

46 New Uses of Micro and Nanomaterials

*Ds* <sup>=</sup> [ *<sup>d</sup>*(*<sup>t</sup>* ) ]4 *kT* \_\_\_\_\_\_\_

XRD data were adopted for calculation of *D*<sup>s</sup>

ity, *D*<sup>s</sup>

dealloyed Ti60Cu39Au<sup>1</sup>

Dan et al. [27] Copyright Elsevier.

**Figure 6.** Mean pore size (a), surface diffusivity (b), and refining factor of dealloyed Ti60Cu39M<sup>1</sup> . M is the selected LSD elements including Ag, Au, Ni, Pd, and Pt. The lines are used for view guide of data points in G-I and G-II. Reproduced with permission from Dan et al. [30] Copyright Materials Transactions.

can be modified. As shown in **Figure 5b**, the value of *D*<sup>s</sup> was estimated to be 2.5 × 10−18 m2 s−1 for Ti60Cu40 ribbon, 6.0 × 10−20 m2 s−1 for Ti60Cu40 ribbon, 9.7 × 10−21 m2 s−1 for Ti60Cu39Au<sup>1</sup> ribbon and 2.0 × 10−21 m2 s−1 for Ti60Cu39Ni<sup>1</sup> ribbon, 2.3 × 10−22 m2 s−1 for Ti60Cu39Pd<sup>1</sup> ribbon, and 1.3 × 10−22 m2 s−1 for Ti60Cu39Pt<sup>1</sup> ribbon, respectively. The surface diffusivity decreased more than four orders of magnitude due to the addition of Pd and Pt. Compared with *D*<sup>s</sup> of the Au-group-stabilized NPCs, the decrease in *D*<sup>s</sup> in the Pt-group-stabilized NPCs was remarkable.

The refining factor, *R*, is defined as the ratio between the surface diffusivity, *D*<sup>s</sup> , of NPC obtained from amorphous Ti60Cu40 precursor (numerator) and NPCs stabilized by LSDs (denominator):

 *<sup>R</sup>* <sup>=</sup> *Ds Cu* \_\_\_ *Ds <sup>L</sup>* <sup>≈</sup> [ *<sup>d</sup>* (*<sup>t</sup>* )*Cu* ]4 \_\_\_\_\_\_\_ [ *<sup>d</sup>* (*<sup>t</sup>* )*<sup>L</sup>* ]4 <sup>×</sup> [ *aCu* ]4 \_\_\_\_\_ [ *<sup>a</sup> <sup>L</sup>* ]4 (2)

As shown in **Figure 6c**, the estimated refining factor was 968 for Ti60Cu39Pd<sup>1</sup> ribbon and 1780 for Ti60Cu39Pt<sup>1</sup> ribbon, respectively. In other cases, the value of *R* changed from 4 for the dealloyed Ti60Cu39Ag<sup>1</sup> ribbon, 111 for the dealloyed Ti60Cu39Ni<sup>1</sup> ribbon, and 231 for the dealloyed Ti60Cu39Au<sup>1</sup> ribbon, respectively. The refining efficiency of the micro-alloying of the Pt-group elements into the precursor Ti60Cu40 alloy was one order higher than that of the Au-group cases. The lower value of *R* for Ti60Cu40 alloy stabilized by Au-group (G-I) elements could take into account from several aspects: (1) the larger self-diffusion coefficients of Ag and Au than that of Cu [44–48] and (2) the large misfit in the atomic radii [49]. As has been reported before [44–48], the self-diffusion coefficients of Au- and Pt-group elements are ranked as following: Ag > Cu > Au > Ni > Pd > Pt. As such, there are more possibilities for Ag adatoms to meet with other Ag adatoms during the diffusion and rearrangements processes of dealloying to form clusters and to develop the Ag phase because it has a larger self-diffusion coefficient than Cu atoms [44, 48]. The self-diffusion coefficient of Au is slightly smaller than that of Cu [45]. Because the misfit in the atomic radius between Cu and Ag and Cu and Au is about 13%, the incorporation of Ag-Au atoms into the Cu lattice becomes more difficult than that of Cu/Cu atoms. On the other hand, the self-diffusion coefficients of Pt-group (G-II) elements are more than two orders lower than that of Cu element [47, 48], and the formation of the crystalline phase of G-II elements (Ni, Pd, Pt) was interrupted by the fast diffusion of Cu adatoms. Furthermore, the misfits in the atomic radius between Cu and Pd and Pt elements are ca. 8%, and the incorporation of Pd/Pt into the Cu lattices appears more easily. The changes in the lattice constants of the NPCs stabilized by G-I and G-II indicated by XRD patterns in **Figure 1** and Refs. [27, 28] support the present hypothesis. As has presented above, the formation of the Ag and Au crystalline phases after dealloying is considered to be due to the difference in self-diffusion behaviors and misfits of the atomic radius. The minor addition of Group-I LSDs causes the formation of the crystalline Ag/Au phases, and the refining factors of Group I were small. On the other hand, the micro-alloying by Group-II LSDs mainly resulted in the invasion of the Cu lattices, with some Cu atoms in the lattice substituted by Ni, Pd, and Pt atoms, forming solid solutions, and the refining factors for Group II were high. In some extent, these LSD atoms in the lattice are considered to be the main contributors for the elaboration of NPCs.

has been reported [11, 12, 14, 20, 41]. Amorphous precursor alloys with homogeneously distributed Cu, Ti, and LSD atoms also have an important refining effect. The uniform distribution of added Pd and Pt atoms in final NPCs resulted in smaller nanopores, as shown in **Figures 3** and **4**. In 2008, fine nanoporous AuPt alloys with a pore size of about 5 nm were

Elaboration of Nanoporous Copper via Chemical Composition Design of Amorphous Precursor Alloys

effectively reduced the pore size from 10 to 20 nm to about 4 nm, which also supports the present results. However, the high cost of Au and Pt weakens their potential application. By minor addition of the 1 at% Pt-group elements, it is possible to elaborate NPCs down to a pore size of approximately 6 nm, comparable to high-cost AuPt nanoporous in nanoporosity. So far the LSD-stabilized NPCs had a relative small nanopores and narrow ligaments comparable to these of NP Au, NP Pd, and NP Pt. If catalytic Au, Pd, or Pt monolayer is electrodeposited on ultrafine NPCs, the catalytic performance of this kind of cost-effective porous materials is able to be close to those Au, Pd, or Pt catalysts. Meanwhile, if the oxides are assembled on ultrafine NPC templates, the new-developed nanocomposites can be used as promising pseudocapaci-

The 1 at% minor addition of low surface diffusive (LSD) elements in two groups (G-I, Au-group (Ag, Au), and G-II, Pt-group (Ni, Pd, Pt)) is able to elaborate the nanoporous Cu structure dealloyed from micro-alloyed Ti60Cu40 alloy efficiently. The chemical compositions of Ti60Cu40 alloys stabilized by the addition of G-II metals were shown to be more effective in refining nanoporous structure than the addition of G-I metals. Nanoporous Cu with a pore

and Ti60Cu39Pt<sup>1</sup>

solid solutions, such as fcc Cu(Pd) or Cu(Pt) solid solutions, and fcc Cu(Au), Cu(Ag), and Cu(Ni) solid solutions formed after dealloying. The refining factor increases approximately

elaboration was attributed to the dramatic decrease in the surface diffusivity during both preferential dissolution and rearrangement of Cu adatoms. The refinement efficiency of the micro-alloying of the G-II LSDs in Pt-group elements was almost one order higher than that of the G-I LSDs in Au-group elements. The homogeneous distribution of LSD elements in both of the amorphous precursor alloys and the final stabilized NPCs played a key role in refining the NPCs. The strategy outlined in this work has the potential to be applied to other alloy systems to obtain other ultrafine nanoporous metals with comparable nanoporosity to those

The authors gratefully acknowledge the financial support from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) Japan through Grant-In-Aid for Science Research in a Priority Area on "Research and Development Project on Advanced Materials Development and Integration of Novel Structured Metallic and Inorganic Materials" and a

precursor alloy to 1780 for the Ti60Cu39Pt<sup>1</sup>

precursor [24]. The alloying of 6 at% Pt into the Ag65Au35 alloy

http://dx.doi.org/10.5772/intechopen.77222

49

and Ti60Cu39Pt<sup>1</sup>

ribbons after dealloying.

precursor alloy. The

ribbons was confirmed to be bimetallic

fabricated from an Ag65Au29Pt6

**4. Conclusions**

tors to contribute in the energy conversion fields.

size of less than 7 nm was obtained from Ti60Cu39Pd<sup>1</sup>

The residue of dealloyed Ti60Cu39Pd<sup>1</sup>

from 3.7 for the Ti60Cu39Ag<sup>1</sup>

high-cost catalysts.

**Acknowledgements**

The diffusion in the interfacial regions between Cu, LSD adatoms, and Ti adatoms played a key role in the formation of ultrafine NPCs. For example, Evangelakis investigated the diffusion of Au adatoms on Cu and self-diffusion of Cu adatoms by using the molecular dynamic method [50]. They have found that diffusion of Cu adatoms takes place exclusively by hopping from one adatom position to an adjacent one and that multiple jumps are frequent at low temperatures. On the other hand, Au adatoms on the NPC ligaments hopped less frequently than in the case of Cu self-diffusion. The migration energy required for hopping of Au on Cu was almost twice that of the corresponding energy for Cu adatoms. While the diffusion of Au on Cu was more difficult than the diffusion of Cu on Cu, this phenomenon is compatible with the binding energy of Au adatoms, which is found to be 2.77 eV for Au adatoms and 2.26 eV for Cu adatoms. The Au adatoms relaxed at a distance −15% smaller than the bulk interlayer distance, but in the case of the Cu adatoms, it was −10% for the same quantity [50]. The Cu adatoms diffused quickly in and out by themselves; however, the Au adatoms at the activated sites diffused out slowly and hardly ever diffused back to the NPC ligaments. Consequently, Au adatoms gradually accumulated outside the ligaments during the dealloying [27, 30]. As a result, Au adatoms uniformly distributed outside the ligaments after dealloying for 43.2 ks via the hopping mechanism, as shown in **Figure 5**. When more Au was added to TiCu amorphous alloys, the surface coverage of Au adatoms increased during dealloying, suggesting that the accumulation rate of Au adatoms in the more concentrated solutions was higher due to the faster migration of Au and Cu adatoms [27]. Because the accumulated Au adatoms built up a continuous outmost diffusion barrier, the behaviors of diffusion and rearrangements of Cu and Au adatoms can be fulfilled to form ultrafine NPCs from both the LSD-substituted Ti60Cu40 amorphous precursor alloys in HF solutions. The diffusion behavior for other LSDstabilized nanoporous structures was considered to be similar. However, the accumulation of G-II Pt-group elements (Ni, Pd, and Pt) with much slower diffusion rates and Ni, Pd, and Pt cluster or grains with much smaller size less than 7 nm tended to form, which are not sensitive to X-ray [51]. As has been reported [52], a bimodal nanoporous structure with a pore size of 10 nm and 20 nm has been fabricated from Al75Pd17.5Au7.5 precursor alloys by successive dealloying. The initial heterogeneous microstructure consisting of Al<sup>2</sup> Au- and Al3 Pd-type intermetallics causes the formation of a bimodal nanoporous structure. The similar evolution of bimodal or multimodal nanoporosity on precursor alloys heterogeneous in microstructure has been reported [11, 12, 14, 20, 41]. Amorphous precursor alloys with homogeneously distributed Cu, Ti, and LSD atoms also have an important refining effect. The uniform distribution of added Pd and Pt atoms in final NPCs resulted in smaller nanopores, as shown in **Figures 3** and **4**. In 2008, fine nanoporous AuPt alloys with a pore size of about 5 nm were fabricated from an Ag65Au29Pt6 precursor [24]. The alloying of 6 at% Pt into the Ag65Au35 alloy effectively reduced the pore size from 10 to 20 nm to about 4 nm, which also supports the present results. However, the high cost of Au and Pt weakens their potential application. By minor addition of the 1 at% Pt-group elements, it is possible to elaborate NPCs down to a pore size of approximately 6 nm, comparable to high-cost AuPt nanoporous in nanoporosity. So far the LSD-stabilized NPCs had a relative small nanopores and narrow ligaments comparable to these of NP Au, NP Pd, and NP Pt. If catalytic Au, Pd, or Pt monolayer is electrodeposited on ultrafine NPCs, the catalytic performance of this kind of cost-effective porous materials is able to be close to those Au, Pd, or Pt catalysts. Meanwhile, if the oxides are assembled on ultrafine NPC templates, the new-developed nanocomposites can be used as promising pseudocapacitors to contribute in the energy conversion fields.
