letters to nature

Bandshift assays. thermophilic archaeon Methanobacterium thermoactotrophicum THF. EMBO J. 15, 5459±5469
(1996).
Standard bandshift reactions (Fig. 3a) utilized 32P-labelled `BH' probes that contained
18. Asahara, H., Wistort, P. M., Bank, J. F., Bakerian, R. H. & Cunningham, R. P. Puri®cation and
three CpG sites (underlined below) whose methylation/mismatch status was varied. The
characterization of Escherichia coli endonuclease III from the cloned nth gene. Biochemistry 28, 444±
BH probes used in Fig. 3c were
4449 (1989).
59TCAGATTCGCGCZGGCTGCGATAAGCTGZGCGGATCCZGGGAATTCAGCT39 19. Shiota, S. & Nakayama, H. UV endonuclease of Micrococcus luteus, a cyclobutane pyrimidine dimer-
DNA glycosylase/abasic lyase: cloning and characterization of the gene. Proc. Natl Acad. Sci. USA 94,
39AGTCTAAGCGCGGYCGACGCTATTCGACGYGCCTAGGGYCCTTAAGTCGA59 593±598 (1997).
20. Robson, C. N. & Hickson, I. D. Isolation of cDNA clones encoding a human apurinic/apyrimidinic
where Y ˆ C, m5C, T or U, and Z ˆ C or m5C (see underlined dinucleotides). The three endonuclease that corrects DNA repair and mutagenesis defects in E. coli xth (exonuclease III)
CpG sites were identically modi®ed/mismatched within each probe. Binding reactions mutants. Nucleic Acids Res. 19, 5519±5523 (1991).
were carried out at room temperature for 30 min in 20 mM HEPES pH 7.9, 25 mM NaCl, 21. Scharer, O. D., Nash, H. M., Jiricny, J., Laval, J. & Verdine, G. L. Speci®c binding of a designed
10 mM b-mercaptoethanol, 1 mM EDTA, 4% glycerol, 1% digitonin and 50 ng sonicated pyrrolidine abasic site analog to multiple DNA glycosylases. J. Biol. Chem. 273, 8592±8597 (1998).
E. coli DNA. The glycosylase reaction does not occur under these conditions (data not 22. Gallinari, P. & Jiricny, J. A new class of uracil-DNA glycosylases related to human thymine-DNA
shown). Complexes were electrophoresed through 6% polyacrylamide gels in 0:5 3 TBE glycosylase. Nature 383, 735±738 (1996).
23. Sibghat-Ullah et al. Base analog and neighboring base effects on substrate speci®city of recombinant
at 4 8C.
human G:T mismatch-speci®c thymine DNA glycosylase. Biochemistry 35, 12926±12932 (1996).
Complexes formed under conditions favourable to the glycosylase reaction (Fig. 3b, c)
24. Bellacosa, A. et al. MED1, a novel human methyl-CpG-binding endonuclease, interacts with DNA
utilized the ¯uorescent or 32P-labelled JJ oligonucleotide (see above) as a probe. In
mismatch repair protein MLH1. Proc. Natl Acad. Sci. USA 96, 3969±3974 (1999).
standard gel-retardation reactions, 200 nM protein was incubated with 66 nM labelled
25. Modrich, P. & Lahue, R. Mismatch repair in replication ®delity, genetic recombination and cancer
oligonucleotide substrate and 333 nM unlabelled homoduplex oligonucleotide in 50 mM
biology. Annu. Rev. Biochem. 65, 101±133 (1996).
Tris±HCl pH 8.0, 1 mM DTT, 5% glycerol, 1 mM EDTA at 37 8C for 20 min. The samples
26. Jiricny, J. Replication errors: cha(lle)nging the genome. EMBO J. 17, 6427±6436 (1998).
were electrophoresed immediately through a 6% native 0:5 3 TBE polyacrylamide gel for
27. Maniatis, T., Fritsch, E. F. & Sambrook, J. Molecular Cloning: A Laboratory Manual 196±198 (Cold
45 min at 100 V. A probe with an abasic site was generated by treatment of oligonucleotide Spring Harbor, New York, 1982).
JJ containing a MG×UG mismatch with the enzyme uracil DNA glycosylase.

Received 25 June; accepted 21 July 1999. Acknowledgements

1. Duncan, B. K. & Miller, J. H. Mutagenic deamination of cytosine residues in DNA. Nature 287, 560± We thank P. Hunziger for HPLC analysis; P. SchaÈr, R. Wood and T. Lindahl for fruitful
561 (1980). discussions; I. Hickson for the HAP1 endonuclease; and S. Tweedie for comments on the
2. Britten, R. J., Baron, W. F., Stout, D. B. & Davidson, E. H. Sources and evolution of human Alu manuscript. This work was supported by grants from the Wellcome Trust (B.H. and A.B.),
repeated sequences. Proc. Natl Acad. Sci. USA 85, 4770±4774 (1988). the Schweizerische Krebsliga (U.H. and J.J.) and by a Darwin Trust Scholarship to H-H.N.
3. Sved, J. & Bird, A. The expected equilibrium of the CpG dinucleotide in vertebrate genomes under a
mutation model. Proc. Natl Acad. Sci. USA 87, 4692±4696 (1990). Correspondence and requests for materials should be addressed to A.B. (e-mail:
4. Bulmer, M. Neighboring base effects on substitution rates in pseudogenes. Mol. Biol. Evol. 3, 322±329 A.Bird@ed.ac.uk).
(1986).
5. Bird, A. P. DNA methylation and the frequency of CpG in animal DNA. Nucleic Acids Res. 8, 1499±
1594 (1980).
6. Cooper, D. N. & Youssou®an, H. The CpG dinucleotide and human genetic disease. Hum. Genet. 78,
151±15 (1988). .................................................................
7. Hollstein, M. et al. Database of p53 gene somatic mutations in human tumors and cell lines. Nucleic
Acids Res. 22, 3551±3555 (1994). erratum
8. Jones, P. A., Rideout, W. M., Shen, J.-C., Spruck, C. H. & Tsai, Y. C. Methylation, mutation and cancer.
BioEssays 14, 33±36 (1992).
9. Lindahl, T. An N-glycosidase from Escherichia coli that releases free uracil from DNA containing
deaminated cytosine residues. Proc. Natl Acad. Sci. USA 71, 3649±3653 (1974).
10. Lindahl, T., Karran, P. & Wood, R. D. DNA excision repair pathways. Curr. Opin. Genet. Dev. 7, 158± Full sintering of
powdered-metal bodies
169 (1997).
11. Wiebauer, K. & Jiricny, J. In vitro correction of G×T mispairs to G×C pairs in nuclear extracts from
human cells. Nature 339, 234±236 (1989).
12. Neddermann, P. et al. Cloning and expression of human G/T mismatch-speci®c thymine-DNA
glycosylase. J. Biol. Chem. 271, 12767±12774 (1996).
in a microwave ®eld
13. Nan, X., Meehan, R. R. & Bird, A. Dissection of the methyl-CpG binding domain from the
chromosomal protein MeCP2. Nucleic Acids Res. 21, 4886±4892 (1993). Rustum Roy, Dinesh Agrawal, Jiping Cheng
14. Hendrich, B. & Bird, A. Identi®cation and characterization of a family of mammalian methyl-CpG & Shalva Gedevanishvili
binding proteins. Mol. Cell. Biol. 18, 6538±6547 (1998).
15. Nan, X., Campoy, J. & Bird, A. MeCP2 is a transcriptional repressor with abundant binding sites in
genomic chromatin. Cell 88, 471±481 (1997). Nature 399, 668±670 (1999)
...................................................................................................................................
16. Michaels, M. L., Pham, L., Nghiem, M., Cruz, C. & Miller, J. H. MutY, an adenine glycosylase active on
In the penultimate paragraph of this Letter, there are two instances
G-A mispairs, has homology to endonuclease III. Nucleic Acids Res. 18, 3841±3845 (1990).
17. Horst, J. P. & Fritz, H. J. Counteracting the mutagenic effect of hydrolytic deamination of DNA where ``electric ®eld'' has been used when ``magnetic ®eld'' is meant:
5-methylcytosine residues at high temperature: DNA mismatch N-glycosylase Mig.Mth of the one is in the fourth sentence and the other in the last. M

304
© 1999 Macmillan Magazines Ltd NATURE | VOL 401 | 16 SEPTEMBER 1999 | www.nature.com
letters to nature
well known average velocity (v ¼ 0:89 nm s 2 1 ) and orientation. 30. Bauer, P. et al. Depth sensitive visualisation of irradiation-induced columnar defects in 2H-NbSE2.
Euro. Phys. Lett. 23, 585–591 (1993).
Taking the derivative of these trajectories yields a plot of the
longitudinal velocities versus time (Fig. 2b). For clarity, we shifted Supplementary information is available on Nature’s World-Wide Web site (http://www.nature.com).

the results for vortex A and C by 61 nm s−1. The error bar denotes Acknowledgements. We thank A. P. Volodin for developing the first version of our STM, R. J. Drost for
the uncertainty in v (dv ¼ 6 0:15 nm s 2 1 ) due to the error in the auxiliary experiments, M. Konczykowski for carrying out the heavy ion irradiation at GANIL (Caen,
France), and M. V. Marchevsky, P. Le Doussal and T. Giamarchi for discussions. This project was partially
vortex positions. The horizontal bar marks the time needed to supported by FOM and the Dutch-Russian Science Collaboration financed by NWO.
traverse a distance 10a0 (630 nm). It is clearly seen that the velocity
Correspondence and requests for materials should be addressed to P.H.K. (e-mail: kes@phys.leidenuniv.nl).
is modulated with a periodic component with amplitude Dv<
0:6 nm s 2 1 and period a0/v. In the Fourier transform (Fig. 2c), this
shows up as a peak at 14.1 mHz. Such a peak at the ‘‘washboard’’
frequency is known for periodic structures driven through random
point disorder9, but to the best of our knowledge this is the first Full sintering of
time that this washboard feature has been found to survive in the
creep regime. The periodic modulation also appears in the longi- powdered-metal bodies
tudinal displacement determined with respect to the average,
uniform motion versus time. However, it is not observed in a in a microwave field
plot of the transverse displacements versus time. Rather, we see on
our timescale a random-walk-like transverse excursion of the Rustum Roy, Dinesh Agrawal, Jiping Cheng
bundle up to an amplitude 0.2a0, not yet enough to decide how & Shalva Gedevanishvili
the mean-squared displacement grows with time. We leave this for Materials Research Laboratory, The Pennsylvania State University,
future investigations. University Park, Pennsylvania 16802, USA
Fast imaging of vortex lattices by STM provides the possibility of .........................................................................................................................

studying the collective and plastic flow behaviour of elastic media The use of microwaves to process absorbing materials was studied
through various configurations of disorder. M intensively in the 1970s and 1980s, and has now been applied to a
Received 7 December 1998; accepted 5 April 1999. wide variety of materials1–4. Initially, success in microwave heat-
1. Blatter, G. et al. Vortices in high-temperature superconductors. Rev. Mod. Phys. 66, 1125–1388 (1994).
ing and sintering was confined mainly to oxide and some non-
2. Koshelev, A. E. & Vinokur, V. M. Dynamic melting of the vortex lattice. Phys. Rev. Lett. 73, 3580–3583 oxide ceramics5–11; but recently the technique has been extended
(1994). to carbide semimetals12–14 used in cutting tools. Here we describe
3. Giamarchi, T. & LeDoussal, P. Moving glass phase of driven lattices. Phys. Rev. Lett. 76, 3408–3411
(1996). the microwave sintering of powdered metals to full density. We are
4. Le Doussal, P. & Giamarchi, T. Moving glass theory of driven lattices with disorder. Phys. Rev. B 57, able to sinter a wide range of standard powdered metals from
11356–11403 (1998).
5. Moon, K., Scalettar, R. & Zimámny, G. T. Dynamical phases of driven vortex systems. Phys. Rev. Lett.
commercial sources using a 2.45-GHz microwave field, yielding
77, 2778–2781 (1996). dense products with better mechanical properties than those
6. Scheidl, S. & Vinokur, V. M. Driven dynamics of periodic elastic media in disorder. Phys. Rev. B 57, obtained by conventional heating. These findings are surprising
13800–13810 (1998).
7. Balents, L., Marchetti, M. C. & Radzikovsky, L. Nonequilibrium steady states of driven periodic in view of the reflectivity of bulk metals at microwave frequencies.
media. Phys. Rev. B 57, 7705–7739 (1998). The ability to sinter metals with microwaves should assist in the
8. Nattermann, T. Scaling approach to pinning; Charge-density waves and giant flux creep in super-
conductors. Phys. Rev. Lett. 64, 2454–2457 (1990).
preparation of high-performance metal parts needed in many
9. Olsen, C. J., Reichhardt, C. & Nori, F. Nonequilibrium dynamics phase diagram for vortex lattices. industries, for example, in the automotive industry.
Phys. Rev. Lett. 81, 3757–3760 (1998). Reviews of microwave processing1–5 describe its use for materials
10. Osakabe, N., Kasai, H., Kodama, T. & Tonomura, A. Time-resolved analysis in transmission electron
microscopy and its application to the study of the dynamics of vortices. Phys. Rev. Lett. 78, 1711–1714 ranging from wood, bacon and potato chips to rubber, ceramics and
(1997). semiconductors, but make no mention of metal sintering. Walk-
11. Kirtley, J. R. et al. Direct imaging of integer and half-integer Josephson vortices in high-Tc grain
boundaries. Phys. Rev. Lett. 76, 1336–1339 (1996).
iewicz et al.15 exposed a range of materials, including six metals
12. Oral, A. et al. Direct observation of melting of the vortex solid in Bi2Sr2CaCu2O8+d single crystals. Phys. (presumably partly oxidized in air) to a 2.4-GHz field, and reported
Rev. Lett. 80, 3610–3613 (1998). modest heating (but not sintering), ranging from 120 8C for Mg to
13. Mozer, A. et al. Observation of single vortices condensed into a vortex-glass phase by magnetic force
microscopy. Phys. Rev. Lett. 74, 1847–1850 (1995). 768 8C for Fe. Sintering of tungsten carbide–cobalt composites has
14. Bolle, C. A., de la Cruz, F., Gammel, P. L., Waszczak, J. V. & Bishop, D. J. Observation of tilt induced also been reported16. Nishitani reported17 that by adding a few per
orientational order in the magnetic flux lattice 2H-NbSe2. Phys. Rev. Lett. 71, 4039–4042 (1993).
15. Yao, Z. et al. Path of magnetic flux lines through high-Tc copper oxide superconductors. Nature 371,
cent of electrically conducting powders such as aluminium, the
777–779 (1994). heating rates of refractory ceramics was considerably enhanced; but
16. Marchevsky, M. V., Aarts, J., Kes, P. H. & Indenbom, M. V. Observation of the correlated vortex flow in no mention was made of the microwave sintering of pure metal
NbSe2. Phys. Rev. Lett. 78, 531–534 (1997).
17. Pardo, F. et al. Real space images of the vortex lattice structure in a Type II superconductor during powders. Whittaker and Mingos18 used the highly exothermic (and
creep over a barrier. Phys. Rev. Lett. 79, 1369–1372 (1997). rapid) reactions of metal powders with sulphur for the microwave-
18. Hess, H. F., Robinson, R. B. & Waszczak, J. V. Vortex-core structure observed with a scanning
tunneling microscope. Phys. Rev. Lett. 64, 2711–2714 (1990).
induced synthesis of metal sulphides. German19 discusses micro-
19. Maggio-Aprile, I., Renner, Ch., Erb, A., Walker, E. & Fischer, Ø. Direct vortex lattice imaging and wave heating of oxide and non-oxide ceramics, but does not
tunneling spectroscopy of flux lines on Yba2Cu3O7−d. Phys. Rev. Lett. 75, 2754–2757 (1995). mention sintering of metals.
20. Renner, Ch., Revaz, B., Kadowaki, K., Maggio-Aprile, I. & Fischer, Ø. Observation of the low
temperature pseudogap in the vortex cores of Bi2Sr2CaCu2O8+d. Phys. Rev. Lett. 80, 3606–3609 (1998). We have used commercial powdered metal components of
21. De Wilde, Y. et al. Scanning tunneling microscopy observation of a square Abrikosov lattice in various alloy compositions—including iron and steel, copper,
LuNi2B2C. Phys. Rev. Lett. 78, 4273–4276 (1997).
22. Volodin, A. P., Golubov, A. A. & Aarts, J. Vortex core shapes measured by STM. Z. Phys. B 102, 317–
aluminium, nickel, molybdenum, cobalt, tungsten, tungsten car-
321 (1997). bide and tin, and their alloys20 —to obtain essentially fully dense
23. Volodin, A. P. & Troyanovski, A. M. A three-phase piezoinertio motor for a low temperature STM. bodies with substantially improved mechanical properties com-
Instr. Exp. Techn. 40, 724–726 (1997).
24. Wengel, C. & Tauber, U. C. Weakly pinned Bose glass vs Mott insulator phase in superconductors. pared to similar bodies sintered by conventional thermal means.
Phys. Rev. Lett. 78, 4845–4848 (1997). Some of these samples were obtained from Keystone Powdered
25. Behler, S. et al. Vortex pinning in ion-irradiated NbSe2 studied by scanning tunneling microscopy.
Phys. Rv. Lett. 72, 1750–1753 (1994).
Metal Company (St. Mary, Pennsylvania, USA) and some were
26. van der Beek, C. J., Nieuwenhuys, G. J. & Kes, P. H. Nonlinear current diffusion in type-II made in our laboratory. They were all standard ‘green’ parts—metal
superconductors. Physica C 197, 320–336 (1992). powders cold-pressed with a few per cent of organic binder. The
27. Larkin, A. I. & Ovchinnikov, Yu. N. Pinning in type II superconductors. J. Low Temp. Phys. 34, 409–
427 (1979). typical size range was 2–5 cm in the largest dimension, with a wide
28. Angurel, L. A., Amin, F., Polichetti, M., Aarts, J. & Kes, P. H. Dimensionality of collective pinning in variety of rectangular and cylindrical cross-sections, including
2H-NbSe2 single crystals. Phys. Rev. B 56, 3425–3432 (1997).
29. Schmid, A. & Hauger, W. On the theory of vortex motion in an inhomogeneous superconducting film.
typical toothed gears in the 1–4-cm range.
J. Low Temp. Phys. 11, 667–685 (1973). Over the past decade we have developed several specialized

668
© 1999 Macmillan Magazines Ltd NATURE | VOL 399 | 17 JUNE 1999 | www.nature.com
 letters to nature
microwave sintering chambers capable of processing a variety of Table 1 Properties of microwave and conventionally processed powdered-
samples with different shapes and sizes, and maintaining any metal samples
temperature in the range from room temperature to 2,000 8C. The Sample Process Green Sintered Rockwell MOR
microwave generators are operated at a frequency of 2.45 GHz, with density density hardness (103 lb in−2)
(g cm−3) (g cm−3)
power output in the range 1–6 kW, in both single- and multi-mode .............................................................................................................................................................................
operation. Inside a typical microwave cavity is an alumina tube Fe-Ni MW 7.11 7.15 B82 177
(industrial part) Conv. 7.10 B77 109
surrounded by ceramic fibre (typically mullite) insulation. The .............................................................................................................................................................................
primary function of insulation is to preserve the heat inside the Fe-Cu MW 6.81 7.17 B96 142
(industrial part) Conv. 6.84 B80 118
tube. Insulation does not absorb microwaves in the lower tempera- .............................................................................................................................................................................
Fe-Cu MW 6.95 6.96 B75 134
ture ranges, but at higher temperatures there is some partition of the (lab. sample) Conv. 6.95 B64 122
power dissipation between the insulation and the sample. For .............................................................................................................................................................................
MW, microwave processed; conv., conventionally processed. The modulus of rupture
higher temperatures (above ,1,400 8C) we use ZrO2 and even (MOR) of most microwave-processed samples is higher than that of the conventional
Y2O3 insulation, but these are much more expensive, and the samples. The densities of many microwave-processed samples are also higher than
those of conventional samples.
partition is of course different. For some runs we utilize a secondary
coupler or microwave susceptor rods made of SiC or MoSi2.
Temperatures are read by optical pyrometers and/or sheathed
thermocouples placed very close to the surface of the sample. experiments by removing the susceptors altogether and monitoring
Although relative readings are accurate to 65 8C, the absolute the power level. It was also established in essentially all cases that the
temperature has not been established to anywhere near that net shape of the green part was retained precisely (with the usual
precision. The atmosphere can be controlled as needed: from air, shrinkage if any was planned), and a fine microsctructure was
to H2, to forming gas, to oxygen. A diagram of a typical microwave produced.
system for sintering of powdered metals is shown in Fig. 1. Table 1 shows data for some of these microwave-processed parts,
The ‘green’ commercial powdered-metal bodies are introduced and the corresponding properties of parts of the same composition
into the microwave chamber and heated at 1,100–1,300 8C typically made by conventional means. Our findings indicate that every
for times ranging from 5 minutes to one hour in flowing forming powdered-metal ‘green’ body so far tested could be sintered in
gas (N2 þ H2 ) or in a pure hydrogen atmosphere. In a large number 10–30 minutes in an appropriate microwave sintering apparatus. It
of cases, similar samples have been heated in conventional furnaces is also clear that in almost all cases, the modulus of rupture (MOR)
for direct comparison of the properties. The first few runs produced of microwave-processed samples is substantially higher than the
highly sintered bodies in a very short period of time. This was conventional samples; in the case of the Fe–Ni composition, it was
achieved in our controlled-atmosphere microwave system with 60% higher. The densities of many microwave-processed samples
2.45-GHz frequency and a maximum 6-kW power: but only are also higher than conventional samples.
1.4 kW of power was used to attain the desired sintering tempera- The samples with a composition of Fe þ Cuð2%Þ þ graphite
tures. Typically the total cycle time was ,90 minutes. For this ð0:8%Þ were processed in a microwave field at 1,200 8C for 30
sintering process, we have established that similar heating occurs minutes. The microwave-sintered and ‘green’ samples were char-
without using susceptors. The only difference is that with susceptors acterized for their microstructure by optical microscopy and for
the overall heating is faster. This was proved by conducting several phase composition by X-ray diffractrometry. The microstructure

Microwave cavity a b
Fe-Cu-C system (green) Fe-Cu-C system (sintered)

Copper

Insulation

Gas
Ceramic 50 µm 50 µm
Sample tube
c d
Fe-Ni-C system (green) Fe-Ni-C system (sintered)
SiC/MoSi2
susceptor Nickel

Microwaves
50 µm 50 µm

Figure 1 Diagram of microwave system for sintering of powdered-metal parts. Figure 2 Optical micrographs of two sets of ‘green’ (that is, unsintered) and
The system consists of a 2.54-GHz microwave oven (cavity) and a ceramic microwave-sintered powdered–metal parts processed at 1,200 8C for 30 minutes.
(alumina) tube which is inserted into the cavity and surrounded by ceramic fibre a, b, Fe-Cu-C; ‘green’ sample (a), microwave-sintered sample (b). c, d, Fe-Ni-C;
blocks. Inside the insulation, SiC or MoSi2 susceptor rods are inserted. The ‘green’ sample (c), microwave-sintered sample (d). Microwave-sintered
sample is placed inside the ceramic tube. The system is capable of achieving powdered-metal parts show excellent sintering of Fe particles. In the Fe-Cu-C
temperatures up to 1,600 8C, and any desired atmospheres (such as H2, N2, Ar) system, copper melted and spread into Fe particle boundaries forming Fe-Cu
can be used. The same system has been used with or without the susceptor (SiC/ solid solutions (b). In the Fe-Ni-C system, the microstructure shows light shades
MoSi2) rods; sintering times without the rods are somewhat longer but no other for austenitic nickel-rich islands (d). In both cases, the formation of steel was
change is noted. confirmed by X-ray diffraction.

NATURE | VOL 399 | 17 JUNE 1999 | www.nature.com

© 1999 Macmillan Magazines Ltd 669
letters to nature
photographs (Fig. 2) indicate that excellent sintering had occurred Another possible significant contribution to the absorption of
between the iron particles, and that the copper had melted and microwave power, on which much theoretical work has been done22,
spread into the iron-particle boundaries forming Fe–Cu solid is multiple scattering in powdered assemblages; this is especially
solutions. The X-ray diffractograms confirm the microscopy obser- applicable to ceramics and higher microwave frequencies. M
vations, and show that the sintered sample contained only a a-iron Received 9 October 1998; accepted 20 April 1999.
solid solution single phase.
1. Clark, D. & Sutton, W. H. Microwave processing of materials. Annu. Rev. Mater. Sci. 26, 299–331
We also microwave-sintered several powders of pure metals. Only (1996).
one representative example is described here. Cobalt metal 2. Schiffman, R. F. Commercializing microwave systems: Paths to success or failure. Ceram. Trans. 59, 7–
17 (1995).
powder was pressed into pellets, and microwave sintered in pure 3. Katz, J. D. Microwave sintering of ceramics. Annu. Rev. Mater. Sci. 22, 153–170 (1992).
H2 at 1 atmosphere pressure at various temperatures ranging from 4. Sutton, W. Microwave processing of ceramics: an overview. Mater. Res. Soc. Symp. Proc. 269, 3–19
900 to 1,200 8C for 10 minutes. The density reached 8.70 g cm−3 at (1992).

900 8C to 8.88 g cm−3 at 1,000–1,050 8C, and very near theoretical

5. Sutton, W. Microwave processing of ceramic materials. Am. Ceram. Soc. Bull. 68, 376–386 (1989).
6. Fang, Y., Agrawal, D. K., Roy, D. M. & Roy, R. Fabrication of transparent hydroxyapatite ceramics by
density of 8.89 g cm−3 at 1,100–1,200 8C. microwave processing. Mater. Lett. 23, 147–151 (1995).
7. Agrawal, D. K., Fang, Y., Roy, D. M. & Roy, R. Mater. Res. Soc. Symp. Proc. 269, 231–236 (1992).
These findings have implications for the field of powdered-metal 8. Fang, Y., Agrawal, D. K., Roy, D. M. & Roy, R. Fabrication of porous hydroxyapatite ceramics by
technology. Metal powders are used in industry for a wide range of microwave processing. J. Mater. Res. 7, 490–494 (1992).
products and applications. Traditional powder metallurgy involves 9. Fang, Y., Roy, R., Agrawal, D. K. & Roy, D. M. Transparent mullite ceramics from diphasic aerogels by
microwave and conventional processing. Mater. Lett. 28, 11–15 (1996).
a process whereby a metal or alloy powder is compacted as a ‘green’ 10. Fang, Y., Agrawal, D. K., Roy, D. M. & Roy, R. Microwave sintering of calcium strontium zirconium
body and sintered to net shape at elevated temperatures. This phosphate ceramics. Ceram. Trans. 36, 109 (1993).
11. Fang, Y., Agrawal, D. K., Roy, D. & Roy, R. Microwave sintering of hydroxyapatite-based composites.
process is not labour-intensive, but does consume considerable Ceram. Trans. 36, 397 (1993).
energy. It conserves materials, and produces high-quality compo- 12. Roy, R., Agrawal, D., Cheng, J. P. & Mathis, M. Microwave processing: triumph of applications-driven
nents with reproducible properties. However, the need for materials science in WC-composites and ferroic titanates. Ceram. Trans. 80, 3–26 (1997).
13. Cheng, J. P., Agrawal, D. K., Komarneni, S., Mathis, M. & Roy, R. Microwave processing of WC-Co
of high integrity for advanced engineering applications requires composites and ferroic titanates. Mater. Res. Innov. 1, 44–52 (1997).
newer technologies. The attainment of finer microstructures and 14. Gerdes, T. & Willert-Porada, M. Mater. Res. Soc. Symp. Proc. 347, 531–537 (1996).
15. Walkiewicz, J. W., Kazonich, G. & McGill, S. L. Microwave heating characteristics of selected minerals
near-theoretical densities in special powdered-metal components is and compounds. Min. Metall. Processing 5, 39–42 (1988).
still elusive and challenging. Increasing cost is also a concern. The 16. Willert-Porada, M., Gerdes, T., Rodiger, K. & Kolaska, H. Einsatz von Mikrowellen zum Sintern
microwave processing method reported here may offer fine micro- pulvermetallurgischer Prudukte. Metall 50, 744–752 (1996).
17. Nishitani, T. Method for sintering refractories and an apparatus therefor. US Patent No. 4147911
structures and better properties in powdered-metal products at (1979).
lower cost. The main reason why the microwave process yields 18. Whittaker, A. G. & Mingos, D. M. Microwave-assisted solid-state reactions involving metal powders. J.
Chem. Soc. Dalton Trans. 2073–2079 (1995).
better mechanical properties is two-fold, especially in the case of 19. German, R. M. Sintering Theory and Practice (Wiley, New York, 1996).
powdered metals: it produces finer grain size, and the shape of the 20. Gedevanishvili, S., Agrawal, D. & Roy, R. Microwave combustion synthesis and sintering of
porosity (if any) is different from that generated during conven- intermetallics and alloys. Mater. Sci. Lett. (submitted).
21. Cherradi, A., Desgardin, G., Provost, J. & Raveau, B. Electric magnetic field contributions to the
tional heating. In microwave-processed powdered-metal samples microwave sintering of ceramics. In Electroceramics IV Vol. II, (eds. Wasner, R., Hoffmann, S.,
we observed round-edged porosities, producing higher ductility Bonnenberg, D. & Hoffmann, C.) 1219–1224 (RWTN, Aachen, 1994).
22. Shanker, B. & Lakhtakia, A. Extended Maxwell Garnett formalism for composite adhesives for
and toughness. microwave-assisted adhesion of polymer surfaces. J. Compos. Mater. 27, 1203–1213 (1993).
The implications for the fundamental science of microwave–
Supplementary information is available on Nature’s World-Wide Web site (http://www.nature.com) or
materials interaction are less clear. What is certain from this work as paper copy from the London editorial office of Nature.
and that of Willert-Porada14 on ceramic sintering is that these
Acknowledgements. We thank J. Kosco for providing powdered-metal components. We also thank the
interactions are more complex than hitherto suspected. The follow- approximately 20 corporations that supported the early stages of this work, which is now supported by the
ing factors, at least, contribute significantly to the total microwave Electric Power Research Institute, the Keystone Powder Metal Company and the Ben Franklin Technology
Center.
heating of powdered metals (object shape and size are also well-
known to play a part, but have not been quantified in detail). First, Correspondence and requests for materials should be addressed to D.A. (e-mail: dxa4@psu.edu).
the microwave field generated by a magnetron and the existence of
an automatic cut-off mechanism to avoid burn-out due to reflected
power. Second, the presence of SiC susceptor rods—taking up only
a small fraction of the total space. Third, the existence of a
cylindrical insulation package made of aluminosilicate fibres essen- Laboratory evolution of
tially transparent to 2.45-GHz radiation at room temperature,
surrounding an alumina tube. Last, the size and nature of the peroxide-mediated
sample placed in the centre of the alumina tube (see Fig. 1).
How the microwave energy absorption is divided between the cytochrome P450
susceptors, insulation, and sample is a generic issue for all micro-
wave processing, and we consider it in more detail in Supplemen- hydroxylation
tary Information. This analysis shows that there is no complete
theory to explain even the simplest case of sintering of a ceramic. Hyun Joo, Zhanglin Lin & Frances H. Arnold
Furthermore, Cherradi et al.21 recently claimed that in most ceram- Division of Chemistry and Chemical Engineering 210-41,
ics the dielectric loss mechanism was a minor contribution to the California Institute of Technology, Pasadena, California 91125, USA
power absorbed compared to the induction losses caused by eddy .........................................................................................................................

currents. These authors also attribute the heating—they did no Enzyme-based chemical transformations typically proceed with
sintering—of metals also to eddy-current losses from the electric high selectivity under mild conditions, and are becoming increas-
field. Their evidence, obtained using samples of different size and ingly important in the pharmaceutical and chemical industries.
shape in different orientations, supports the role of such eddy- Cytochrome P450 monooxygenases (P450s) constitute a large
current losses as a major contributor to the heating of metals. family1 of enzymes of particular interest in this regard. Their
However, all their experiments were done in a single-mode cavity biological functions, such as detoxification of xenobiotics and
where the orientation of the electric and magnetic fields can be steroidogenesis2–5, are based on the ability to catalyse the inser-
fixed. All our work was done in multimode cavities where the tion of oxygen into a wide variety of compounds6. Such a catalytic
situation is much less determined; but the role of the electric field transformation might find technological applications in areas
cannot be ignored. ranging from gene therapy and environmental remediation to

670
© 1999 Macmillan Magazines Ltd NATURE | VOL 399 | 17 JUNE 1999 | www.nature.com