Synthetic Quartz with High Ultraviolet Transmission
`
`A. A. Ballman, D. M. Dodd, N. A. Kuebler, R. A. Laudise, D. L. Wood, and D. W. Rudd
`
`Quartz has been synthesized under hydrothermal conditions at rates of 1.78 mm/day in the presence of
`LiNO2 in Ag lined and Ag plated systems, and it is shown that such quartz has optical transmission
`between 1500 A and 3 ;u equal, and in some cases superior to, natural quartz. The uv cutoif at about
`1500 A has been shown to be associated with Fe whose concentration may be reduced by procedures that
`reduce the concentration of charge compensating H*’ in the lattice (growth in LiNO2) and by pro-
`cedures which reduce the concentration of Fe in solutions (inert conditions). Transmission near 3 [J
`is affected by OH which can be reduced by LlNO2.
`
`Introduction
`
`Experimental
`
`The optical absorption of synthetic quartz in the ir
`region has been studied considerably and it has been
`shown that OH is responsible for absorption in the 3-}!
`region.‘‘3 Recently, the optical absorption has been
`shown to correlate with acoustic absorption‘ and
`methods of hydrothermally growing quartz with low
`OH concentrations and high acoustic Q have been re-
`ported.5-6
`It has been shown that quartz may be pre-
`pared from hydrothermal solutions in which the min-
`eralizer is NaOH and the growth rate is greater than
`1.78 mm/day with an OH concentration in the 15-30
`ppm range and a Q as high as 1 X 10“, provided LlNO2
`is present in the growing solution.“ Mechanisms for
`the effect of LiNO2 on OH concentration and Q have
`been proposed.’ Vitreous silica and crystalline quartz
`with good ir, visible, and uv transmission are of great
`technological interest, and the transmission of low OH
`synthetic crystalline quartz, particularly in the uv, has
`not been reported. Earlier experiments have shown
`that deliberate iron doping markedly reduces the trans-
`mission in the optical
`range.3‘1° Large size inert
`(silver) systems capable of quartz growth in the absence
`of iron have recently become available.“'”
`It now is
`possible,
`therefore,
`to investigate the properties of
`quartz grown in the absence of iron. VVe report here
`the results of a study of the optical transmission of
`quartz prepared under conditions known to exclude
`OH and of quartz prepared in the absence of Fe.
`
`The first five authors are with Bell Telephone Laboratories,
`Inc., Murray Hill, New Jersey 07974; D. W. Rudd is with
`Western Electric Company,
`Inc., Merrimack Valley Works,
`North Andover, Massachusetts.
`Received 29 February 1968.
`
`The conditions and equipment used in this work were
`those ordinarily employed in commercial quartz syn-
`thesis in NaOH solutions”-14:
`crystallization tem-
`perature~352°C;
`temperature difference (At) between
`crystallization and growth zones——44°C; degree of fill——
`82%;
`pressure—1570 atm;
`solution——1 M NaOH
`(0.1 M Li+ salts where they were used); nutrient——fus-
`ing quality lascas—Brazilian crystalline quartz;
`seed
`orientation—basal plane, (0001); and rate—1.78 mm/
`day.
`Growth was conducted under three sorts of condi-
`tions.
`
`(1) Directly in a 4340 alloy steel autoclave. Such
`autoclaves have a protective layer of sodium iron
`silicate (acmite) that has been built up during previous
`runs, but the nutrient basket, baffle, seed frame, and
`seed holders are new low carbon steel in each run.
`In
`
`this paper, synthesis under these conditions is termed
`growth in an iron autoclave.
`(2) Conditions as in (1) except that the nutrient
`basket, baffle, seed frame, and seed holders are plated
`with 0.25 mm of Ag.
`In this paper, synthesis under
`these conditions is termed Ag plated.
`(3) Directly in a silver lined tube“ of the sort
`shown in Fig. 1. This tube is essentially a scaled up
`version of Kolb’s silver lined design.” The tube is sup-
`ported by a steel shell that is more than capable of sup-
`porting the pressure differences between the water in
`the autoclave and the hydrothermal solution in the tube
`that take place during the run. The tube is closed
`with a Morey type” closure and silver sheet seal that
`makes it much more convenient to use than the welded
`closure tubes“ previously used for inert high pressure
`hydrothermal conditions. The percent of fill of water
`outside the tube was 83% for the conditions of our ex-
`periments, and this fill provided a satisfactory pressure
`balance. The seed, frame, baffle, seed holders, and
`
`July 1968 / Vol. 7, No. 7 / APPLIED OPTICS 1387
`
`(cid:34)(cid:52)(cid:46)(cid:45)(cid:1)(cid:18)(cid:17)(cid:19)(cid:25)
`ASML 1028
`
`

`
`ppm OH by weight assuming the relationship ppm =
`100 0:, which has previously been used for crystalline
`quartz.“‘
`Examination of Table I shows that the best quality
`with respect to both uv and ir transmission is obtained
`when OH (addition of LiNO2) and Fe are excluded
`(crystallization in Ag tube). A sample prepared in
`this manner (X-246) is slightly more transparent in the
`uv than natural quartz and equivalent
`to natural
`quartz in the ir. Transmission results are nearly as
`good with only a slight reduction in the ir transmission
`when LiNO2 is present and Fe is partly excluded (Ag
`plated conditions, X—261). Somewhat poorer quartz
`results, as judged by the uv transmission, when no at-
`tempt to exclude Fe is made even though LiNO2 is
`present (X-162), and a further decrease in transmission
`both in the ir and uv occurs when LigCO3 is used
`(1802—S) even though iron is partially excluded. Even
`poorer transmission occurs when no attempt to exclude
`either iron or OH is made (X—1). KOH is shown to be
`much inferior to NaOH as a mineralizer since in its
`presence (X257—23) both the ir and uv transmission are
`poor. There seems to be no particular advantage in
`using nutrient prepared from high purity SiCl4 instead
`of melting lascas, as sample (H-2) shows.
`Fe might be included in SiO2 either substitutionally
`or interstitially. Both Fe“ and Fe“ have been ob-
`served in quartz and are thought to be responsible for
`the colors of citrine and an1oethyst.3—1°’17—19
`Since the
`ionic radius of Fe“+ (0.75 A) is a poor fit to the Si”
`(0.42 A) site,
`it probably occurs interstitially while
`Fe3+ (0.64 A) may occur either substitutionally or in-
`terstitially. Thus, interstitial Fe” may charge com-
`pensate substitutional Fe3+ (Ref. 18) at least when the
`iron concentration is appreciable.
`However,
`the iron concentration in the crystals
`examined in this study is probably too low for Fe self-
`compensation to be important.
`Indeed the uv ab-
`sorption which we have observed in the crystals of this
`study is at much shorter wavelengths (<2400 A)
`than
`that seen in the studies of green quartz, citrine, and
`amethyst (>3400 A). Thus, we believe that at
`the
`low concentrations of Fe present in this study the im-
`portant charge compensating species for substitutional
`Fe“ is OH with the positive charge being provided by
`H+ in the lattice and the 0 being the normal lattice
`oxygen.
`
`Consequently, growth conditions which reduce OH in
`the lattice (Li+ salts and particularly LiNO2) will re-
`duce Fe” since its distribution constant will depend
`upon the ease with which it can be charge compensated.
`In addition, growth conditions which reduce the iron
`concentration in solution (Ag tube, Ag plating) will
`also reduce Fe“ in the lattice. Similarly, OH may be
`reduced by procedures that reduce the Fe concentration
`in the system but since other ions such as Al“ (from
`the nutrient) may charge compensate for I-I+, this ef-
`fect is partially obscured.
`Referring to Fig. 2, it appears that there are at least
`three absorption cognponents in the uv:
`one near 2400
`A, one near 1950 A, and one near 1600 A. The one
`
`I524 cm ID CONTAINING VESSEL
`
`NUT
`
`CONTAINING AUTOCLAVE COVER
`
`
`
`THERMOCOUPLE
`WELL
`
`LIP AREA
`0.228 cm
`STERLING
`
`
`
`
`
`LINING
`O.|2?—0.|78cm
`FINE
`SILVER
`
`
`
`295.3 cm
`
`288.3 cm
`
`
`
`I IL43cm
`I]I"!
`Ell? jlll I
`
`
`
`
`Fig. 1.
`
`Cross section of a silver lined tube for hydrothermal
`growth of quartz crystals.
`
`nutrient basket in the silver tube runs were all made of
`silver.
`In this paper, synthesis under these conditions
`is termed Ag tube. Samples were cut parallel to the
`(0001) axis and in such a way that only (0001) growth
`was examined. Spectra were measured with un-
`polarized radiation on a Perkin—Elmer model 21
`spectrophotometer in the 3-): region, 9n a Cary model
`14 spectrophotometer between 2000 A and 2.5 ,4, and
`on a MacPherson model 225 vacuum uv scanning
`monochromator below 2000 A.
`
`Results and Discussion
`
`Figures 2 and 3 show the absorption spectra of the
`samples prepared in this study.
`In Fin‘. 2, the near
`and far uv are shown on two different ordinate scales to
`
`accommodate the greatly increased absorption in the
`far uv. Figure 3 shows the ir absorption in the 3—;i
`region where the absorption is due to OH stretching
`Vibrations. Table I identifies the crystals studied and
`lists them in increasing order of uv absorption as
`measured by their uv cutoff wavelength defined as the
`wavelength at which the extinction coefficient
`at =
`(1/13)
`logm (I0/I) = 1.5 cm“. As can be seen from
`Table 1,
`the sample order as determined by cutoff
`wavelength is similar to that determined by OH con-
`centration. The OI-I concentration was calculated by
`converting the extinction coeflicient at 3500 cm“ to
`
`1388 APPLIED OPTICS / Vol. 7, No.7 / July 1968
`
`

`
`15;-
`
`_
`'2
`cf
`
`0.5-
`
`X-257-23
`
`1.07
`
`
`I802-8
`x—u_-‘,2
`26!
`246
`
`0
`2000
`
`2500
`>07:
`
`3000
`
`
`
`x-257-23
`
`I500
`
`I600
`
`— —
`I800
`L900
`
`I700
`X, 3
`
`I802~8
`X"
`H7
`X-I62
`
`2000
`
`X- 257-23
`
`E"
`
`_ 6
`5
`
`4
`
`2
`
`0
`
`a = (1 /t) logm (Io/I) is
`Fig. 2 Ultraviolet spectra of synthetic and natural quartz grown under various conditions.
`extinction coefficient per cm of sample thickness. Sample descriptions are in Table I.
`
`the
`
`Table I. Properties of Synthetic Quartz Crystals
`
`Cutoff at
`Crystal
`at = 1.5
`
`designation
`(cm‘1, rat)
`OH (ppm)
`Growth conditions
`
`NaOH + LiNO2
`Ag tube
`4
`1460
`X-246
`NaOH -1- LiNO2
`Ag plated
`17
`1465
`X-261
`NaOH + LiNO2
`Ag tube (SiCl4 nutrient)
`4
`1480
`H-2
`-
`—
`4
`1520
`Natural #117
`NaOH + LiNO2
`Fe autoclave
`14
`1630
`X—162
`NaOH + Li2CO3
`Ag plated
`30
`1740
`1802-8
`Na0H
`Fe autoclave
`41
`1740
`X—1
`
`X257—23 KOH 2200 220 Ag tube
`
`
`
`
`has definitely been associated with Fe”
`near 2400
`in vitreous SiO2 D(Ref. 20), although not all absorption
`bands at 2400 A are due to Fe” (Ref. 21).
`It is
`probable that the other two are also associated with this
`ion and that the uv absorption in our samples is due to
`Fe. From the correlation of OH concentration with
`uv cutoff observed in Table I, it is evident that Fe con-
`centration is linked to H+ charge compensation. Fur-
`thermore, from the correlation of uv cutoff with the
`inertness of growth conditions, it is evident that Fe
`
`concentration in the grown crystals is correlated with
`Fe concentration in the growth solution. The es-
`pecially poor uv transmission in X—257—23, even though
`it was grown under inert conditions, suggests that the
`high OH content associated with KOH growth may
`have been due to particulate inclusions“ of H20 that
`cause scattering and decrease the uv transmission.
`Since the lascas nutrient contains some Al“, it is
`possible that some uv absorption is due to this ion and
`further improvements in transmission might be ex-
`
`July 1968 / Vol. 7, No. 7 / APPLIED OPTICS 1389
`
`

`
`previously established relationship between OH and
`acoustic Q, (Ref. 4) one would expect that procedures
`which reduce OH such as the reduction of the Fe“ con-
`centration in the system will result in Q improvements.
`
`Conclusions
`
`Crystalline quartz having a Very high (comparable
`with or slightly better than natural quartz) uv and ir
`transmission can be produced by hydrothermal growth at
`rates up to 1.78 mm/day under conditions which reduce
`the inclusion of Fe and OH in the crystal. Crystalliza-
`tion in an inert (Ag or Ag plated) system in the presence
`of LiNO2 yields such crystals. The distribution con-
`stant of Fe is dependent upon the OH present in the
`crystal so that LiNOx,, which reduces OH greatly reduces
`Fe and improves both the ir and uv transmission.
`Inert conditions further reduce the Fe concentration
`
`in solution and hence in the crystal with further
`improvement in uv transmission.
`
`We would like to thank J. R. Boie for experimental
`assistance, E. D. Kolb for advice concerning the de-
`sign of the Ag lined tube, and D. B. Fraser for advice on
`crystal H-2.
`
`References
`
`5.
`
`599°
`
`1. A. Kats, Phillips Res. Rep. 17, 133, 201 (1962).
`2.
`J. C. King, D. L. \Vood, and D. M. Dodd, Phys. Rev. Lctt.
`4, 500 (1960).
`3. W. A. Senior, Ph. D. Thesis, Kings College, Univ. of
`London (1962).
`4. D. M. Dodd and D. B. Fraser, J. Phys. Chem. Solids 26,
`673 (1965).
`J. C. King, A. A. Ballman, and R. A. Laudise, J. Phys. Chem.
`Solids 23, 1019 (1962).
`6. A. A. Ballman, R. A. Laudise, and D. W. Rudd, Appl.
`Phys. Lett. 8, 53 (1966).
`7. R. A. Laudise, A. A. Ballman, and J. C. King, J. Phys.
`Chem. Solids 26, 1309 (1965).
`A. A. Ballman, Am. Mineral. 46, 439 (1961).
`L. I. Tsinober and L. G. Chentsova, Kristallografiya 4,
`633 (1959).
`10. L. I. Tsinober, L. G. Chentsova, and A. A. Shternberg,
`Growth of Crystals, A. V. Shubnikov and N. N. Sheftal,
`Eds. (Consultants Bureau, Inc., New York, 1959), Vol. 2,
`p. 45.
`11. A. A. Ballman and D. W. Rudd, to be published.
`12. E. D. Kolb, U. S. Pat. 3,271,114 (6 Sept. 1966).
`13. R. A. Laudise and R. A. Sullivan, Chem. Eng. Frog. 55,
`55 (1959); see also R. A. Laudise and J. W. Nielsen,
`Solid State Physics, F. Seitz and D. Turnbull, Eds.
`(Academic Press Inc., New York, 1961), Vol. 12 for dis-
`cussion of equipment, especially Morey closure.
`14. D. W. Rudd and A. A. Ballman, Western Electric Eng. 1,
`3 (1965).
`15. E. D. Kolb, A. S. Coriell, R. A. Laudise, and A. R. Hutsou,
`Materials Res. Bull. 2, 1099 (1967).
`16. D. M. Dodd and D. B. Fraser, Am. Mineral. 52, 149 (1967).
`17. T. I. Barry and VV. J. Moore, Science 144, 3616 (1964).
`18. T. I. Barry, P. McNamara, and W. J. Moore, J. Chem.
`Phys. 42, 2599 (1965).
`19. G. Lehmann and VV. J. Moore, J. Chem. Phys. 44, 1741
`(1966).
`20. V. Garino-Cauina, Rev. Opt. 34, 365 (1955).
`21. N. H. Turner and H. A. Lee, J. Chem. Phys. 43, 1428 (1965).
`
`<jABSORPT|0N
`
`I/I0 THlCKNESS
`
`4ooo
`
`3500
`
`25000
`FREQUENCY. cm"
`
`2500
`
`Infrared absorption spectra of synthetic and natural
`Fig. 3.
`quartz gi'own under various conditions. The individual curves
`were recorded in percent transmission on approximately equal
`thicknesses, but the ordinates have been displaced for clarity.
`Sample descriptions are in Table I.
`
`pected by its exclusion. The results for H-2 where
`nutrient was made from SiCl4, however, indicate that
`this is not so, since essentially no Al“ was present dur-
`ing growth. Chemical analyses for Fe and Al in the
`concentration ranges present in synthetic quartz have
`not yielded useful data on which to base correlations
`between the ion and a crystal property.7 But from the
`
`1390 APPLIED OPTICS / Vol.7, No.7 / July 1968