UNCLASSIFIED

0 :URITY CLASSIFICATION OF THIS PAGE

LinCUMENTATION PA•GE

UNCLASSITlOE

O SECURITY CLASSIFICATION AUT 3 DISTRIBUTION/AVAILABILITY OF REPORT

N &Approved for public releaseVMb DECLASSiFICATION/OOWNGRADI n* EDistribution unlir,,ted

PERFORMING ORGANIZATION REPFRMUMBER( S 5 MONITORING ORGANIZATION REPORT NUMBER(S)

UK AFOSR.TR. 6_ 0 570

_ 8 NAME OF PERFORMING ORGANIZATION OFFICE SYMBOL 7a NAME OF MONITORING ORGANIZATIONS(~rIf app ~ableCorning Glass Uorksl AFOSR/NC

"c. ADDRESS (City. Slate and ZIP Code) 7b ADDRESS (City, Slate and ZIP Code)

Research & Development Division Bldg 410

Corning, NY 14831 Boiling AFB DC 20332-6448

Sa NAME OF FUNDING/SPONSORING Bb OFFICE SYMBOL 9 PROCUREMENT INSTRUMENT IDENTIFICATION NUMBERORGANIZATION (If applicable)

AFOSR/NC I NC F49620-84-C-0098

8c ADDRESS (City. Stale and ZIP Code) 10 SOURCE OF FUNDING NOS

PROGRAM PROJECT TASK WORK UNITBldg 410 ELEMENT NO NO NO NO

Boiling AFB DC 20332-6448

11 TITLE Ilflnude Security Classificotion), Fluor ide Glasses fo 61102F 2303 A3Bulk Optical and Waveguide Applications _

12. PERSONAL AUTHOR(S)

P. A. Tick-13.. TYPE OF REPORT 13b TIME COVERED 114 DATE OF REPOr T(YrMoDay 7 115PAGECOUNTa

Final IFROMI !Ž~4TOl4 Aug,8 Jan 86 1916 SUPPLEMENTARY NOTATION

17 COSATI CODES 18 SUBJECT TERMS IContinue on reverse if necessary and identify by block number;

FIELD GROUP SUB GRA heavy metal fluoride glass

19. ABSTRACT WContinue on reverse i• necessary and identify by block number,,

The goal of this project completed under Contract No.

F49620-83-C-0090 and F49620-84-C-0098 and sponsored by the Air

Force Office of Scientific Research (AFOSR), was to find a widetransmission window fluoride glass which could be compatible with

chemical vapor deposition. The intent of the work was to usethese glasses for bulk IR optics or for ultra low-loss OWG fibers.A new glass family was discovered and explored under this researchprogram. This new fluoride composition field of CdF 2 -LiF-AlF 3 -PbF 2(CLAP) was identified and patented during the study. Potentially,many Air Force requirements both in bulk optics and optical fibersfor infrared transmission can be met using this new glass.

During the course of this study, the CLAP glasses wereidentified, patented, and characterized. CLAP glasses were found

20 0IS, '-,o ,J,'AVAILAS!LITY OF ABSTRACT j21 ABSTRACT SECURITY CLASSIFICATION

UNCLASSIFIED/UNLIMITED E SAM: AS RPT 6_ OTIC USERS

22a NAME OF RESPONSIBLE INDIVIDUAL 22b TELEHONE NUMBER 22c OFFICE SYMBOLfIndlude Irea Code,

DONALD R. L LRICIi 767-4960 INC

DD FORM 1473, 83 APR EDITION OF 1 JAN 73 IS OBSOLETESECURITY CLASSIFICATION OF THIS PAGE

SECURITY CLASSIFICkTION OF THIS PAGE

to have equivalent stability than more commonly used glasses suchas the fluorozirconates. CLAP glasses have a slightly reducedtransmission window, but this trivial disadvantage is more thanmade up for by its considerably better resistence to water attack.

To demonstrate these new glasses as an infrared transmittingmaterial, water had to be removed from our initial compositions.Water has strong IR absorption bands in the 2.5 - 3.5 micronwavelength spectral region of cperation. This was done with aspecial RAP furnace made available at Corning (built entirely withCorning funds) to be used for tais research. This furnace, alongwith other specialized equipment such as glove boxes, mixingfacilities, and general familiarity with fluoride glass, enabledthis research to proceed rapidly to commercialization. Resultsindicated that CClF 3 , CC1 2F 2 , and CC!F 3 were effective in removingwater.

RAP furnace results showed that in small amounts neitheroxygen or chlorine seems detrimental to glass forming. A onepercent concentration of both ap. ars to enhance vitreousstability and oxygen alone improves durability.

In more recent work, the effect of rare earths on attenuationA was studied. Five fluoride salts of rare earths were doped into a

base CLAP glass and the optical spectra was measured at severalconcentration levels. Results showed Eu and Nd impurities wouldbe harmful to ultra low-loss transmission between 2.5 and 3.5microns.

In conclusion, the CLAP system represents a new heavy metalfluoride glass family with good stability and excellent resistanceto water attack. All the components are compatible with chemicalvapor transport. For infrared transmission applications, watercan be removed by RAP melting. Small amounts of chlorine andoxygen enhance vitreous stability and Eu and Nd impurities werefound to have negative effects on ultra low-loss transmission.Small quantities of this glass are available commercially fromCorning Glass Works, Research an6 Development Division.

SECURITY CLASSIFICATION OF THIS PAGE

AFOSR-TR. U6'ý C570Corning Glass W, orksCorning, New York

FTLUtORIDE) GLASSES FOR

13UL1K OPTICAL. AND 1^AVE2UJIDE'

APPLICATIONS

FINAL, REPORT

Corning Glass WorksP. A. ýTick

Contract No. F49620-84-C-0098

Apo

; n

4U

_'LUORIDE CLASSES FOR

BULK OPTICAL AND WAVEGUIDE

APPLICATIONS

FINAL REPORT

Corning Glass WorksP. A. Tick

Contract No. F49620-84-C-0098

* A 17 F o." ~ThIS teehnivl epa• "rt, has beon ravl•owedl g-d is

Dst ribUt On 3 s Un.1Di tted.

ich;QU, U 1ouaacet

Ada D ICAk.cesion frNT7rS CRA~

AU DTICTA

Dstributo ............

Availability Codes

IDist Spca i .I and oS pecial

A FLUORIDE GLASSES FOR BULK OPTICAL AND WAVEGUIDE APPLICATIONS

FINAL REPORT

CONTRACT NUMBER F49620-84-C-0098

INTRODUCTION

SThe overall goal of the research was to develop a wide

transmission window heavy metal fluoride glass (HMFG) which couldeventually be compatible with chemical vapor deposition (CVD)forming. D'ring the first two years of contract work, a newcomposition field, i.e., CLAP (CdF 2 .LiF.AlF 3 .PbF 2 ) glasses, were

identifiee and the basic chemistry and properties of the systemdetermined. Since the intent of the work is to use such glassesfor bulk IR optics or for ultra-low-loss OWG fibers, further

characterization was required.

It remained to establish the utility of these glasses aspractical infrared transmitting materials. This required that a

method for removing water be found and that the effects ofinfrared absorbing impurities be determined. Work in these two

areas constituted much of the recent effort.

"EXPERIMENTAL RESULTS

In order to upgrade the glass quality over that of initial

composition survey work, the purity of the batch components wasimproved. Sources of high-purity PbF 2 , LiF, and CdF 2 were foundand these materials were used in all of the current experiments.

Table 1 shows typical analysis of these materials. Analysis isroutinely performed at Corning by atomic absorption techniques

that routinely characterize impurities in sub-ppm levels.Although a better source for anhydrous AlF 3 is still being sought,a present supplier of other materials expects high-purity AlF 3 tobe available in early 1986.

: " :

i.

-2-

Glass for the composition experiments was prepared by melting

30g batches under dry nitrogen in 35 ml platinum crucibles at

950*C. it was found that by transferring the hot glass to plati-num foil tubes, reheating, and then quenching in hot silicone oil,9 glass bars 1 cm in diameter by 5 cm long, which are essentially

free of crystals, could be obtained. A transmission electromicro-

graph shown in Figure 1 at 40,000x magnification, confirms the

absence of any second phase regions.[.

The glass bars could be annealed and cut up for further

measurements. The details of experimental characterization wereI 1described previously.

Viscosity was measured by Professor James Shelby at Alfred

University using a microbeam bending technique. The temperaturedependence of viscosity is shown in Figure 2. The glass showsnormal Arrhenius behavior over about three orders of maqnitude in

viscosity (109-1012 poise).

In another set of experiments, the effects of chlorine andoxygen were examined by introduction of 29CdF 2 *9LiF.35AIF 3 .29PbF 2 .

Previous analytical work confirmed the presence of ca. 0.5 w/o 0

as an impurity in glasses melted without RAP conditions. This

experiment established the effect prior to our RAP emperiments.Small substitutions, up to about one weight percent, improved

glass quality. Tg was lowered when chlorine was present and

increased with increasing oxygen. Durability improved at higheroxygen concentrations. The multiphonon edge was shifted towardshorter wavelengths with increasing oxygen concentration andunaffected by chlorine. This is shown in Figure 3. These resultsare consistent with previous results for the ZBLA system.

Five Lare-earth elements, Nd, Sm, Pr. Er, and Eu, were dopedinto a glass of composition 29.5CdF 2 .8.5L_ .?4,5A1F 3 .27.5PbF 2 ,

then formed into quenched bars and the transmission spectra•,.*. measured. Figure 4 shows the details of spectral structure

associated with each rare-earth dopant. The plot of absorption

-3-

coefficient for several rare-earth 'ransitions as a function of

concentration (Figure 5) demonstrates Beer-Lambert behavior.

Since the spectral region between 2.6 and 3.5 microns is where OWG

of these glasses will transmit for minimum attenuation, only the

absorption peaks in this vicinity are important. Those are shown

in Tab~e 2. The absorption wavelength, absorption coefficient,

and specific attenuation for each of the transitions involved are

shown. Since it is not known whether the line shape of the decay

of each peak was Gaussian or Lorentzian, the effect of both shapes

at 2.S and 3.5 microns was computed. The proximity of transition

peaks of Nd and Eu make these two rare earths the most serious

obstacles to ultra-low attenuation in these glasses. The attenua-

tion values (dB/km ppmw) tend to be slightly larger than those

reported for fluorozirconate glasses. 2

The construntion of a furnace, capable of reactive atmosphereprocessing (RAP melci~ng) up to 1 kg of glass, was completed inmid- 985. A schematic diagram and a photograph of the apparatus

are shown in Figure 6. RAP experiments were conducted with a 350gbatch of 25.5CdF 2 .8.5Lir.34.5AIF 3 . 27.5PbF2 glass, which was placedin a covered platinum crucible then lowered into the furnaceenclosure. The melt was then heated in flowing nitrogen to 950 0C

by inductively coupling to the platinum crucible. Once attemperature, a RAP gas could be iatroduced dicectly above the melt

surface. After compleling a RAP cycle, the hot glass was trans-ferred to smaller plaý. ium containers where it was remelted and

cast into bars or patties.

Five RAP gases were tested to date. These include CF4 , NF3,CCIF 3, CC1 2 F2 , and CCl 3 F. Relative wate. concantrations in thefinal glasses were monitored by the intensity of the water band at

2920 nm; the results of the preliminary evaluation of these five

gases is summarized in Table 3. The base tine absorptioncoefficient should lie between 1.8 and 8 pm-I as estimated from a

number of 30g melts in which no RAP was done. The :F4 gas seemsto be completely ineffective .. i' e NF, seems to be at best, veryinefficient. Surprisingly, just a high flow of nitrogen itself

--4-

seems to be fairly effective in removing water. All of the

*• chlorine-containing gases appear to be very efficient in removing

the hydroxyl associated with the 2920 nm band. For example,

Figure 7 shows the changes in both the 2920 nm as well as a second

band at 5400 nin as exposure time to relatively low flows of CClF 3

increases. The absorption coefficient decreases by about a factorof two every fifteen minutes. In another experiment, a glass

treated with CCI 2 F2 for one hour at a faster flow rate showed no

evidence of the 2920 nm hydroxyl band.

DISCUSSION

The CLAP system represents a new HMFG family. It is

comparable in stability to fluorozirconates, has a somewhat

smaller transmission window and is considerably more resistant to

water attack. It also provides a system in which each of the

components is compatible with chemical vapor transport.

The emphasis on the more recent work was to assess the

utility of these glasses for both bulk optic; and optical wave-

guide applications. The attenuation of various rare-earth

transitions appear to be slightly larger in these glasses thin

those reported for the fluorozirconates, which is a slight

disadvantage. Eu and Nd appear to be the most serious of the

rare-earth impurities. Preliminary RAP nmelting experiments have

shown that the chlorine-containing gases, CClF 3 , CCl 2 F 2 , and CCI 3 F

are all effective water getters while NF 3 and CF 4 are not. CCl4

would probably also be effective, but it, like CCl 3 F, is a liquid

at room temperature and would require a bubbler, whereas the other

compounds are gases at room temperature. The actual efficiencies

of each of the gases will still have to be determined.

In small amounts of the order of 1%. neither Cl nor oxygen

seems to be detrimental to glass forming. Oxygen improves the

durability but shifts the multiphonon edge towards the visible.

REFERENCES

1. P. A. Tick, D. A. Thompson and F. J. Quan, Annual Report,

Contract F49620-83-C-0090, August 31, 1964.

2. Y. Ohishi, S. Mitachi, S. Shibata and T. Manabe, Jap. J.

Appl. Phys., 20, March 3, 1981, pp. L191-L193.

A

'4

5.4

,.7

, %N

I -6-

TABLE 1

PURITY OF NEW HIGH-QUALITY BATCH MATERIALS FOR CLAP GLASSES

Impurities ppb*Compound Fe Cr Cu Mn Ni Al1

PbF 2 173 79 406 <85 74 <128

LiF 26 39 2816 14 <55 255

CdF 2 (Supplier 1) <155 368 943 <22 <69 <109CdF 2 (Supplier 2) <204 505 1-125 <29 <91 <143

*All values analyzed by Corning Glass Works.

-- -

N ce1 0 4 ) LCr

Dir0 0 0 c

Ua) bO bOflbO b

0 ) Q) (V w) w

,-4 -4 -4 r-4 -

U) r- 1-4 1-4 ~

(d~ bO bO bO bO bOUo u a) a) 0) (V) a)D

A-, .,-

E-' r, r- c' --

~ 0

0)1.

r- - a 0 ) 00 01 C4C1 ,4 q

a- 0) :)t) 0 a

.m' 4- C

0v im N0'~ -4 M- -4

ob

~'-C 0 0Ll C 0V) 0 c o m 0 4 e Lj-.mI- C1 m')L

'-E-

0

4-CL

TABLE 3

EFFECT OF RAP GAS UPON THE WATER BAND AT 2920 nm IN29.5 CdF 2 *8.5 LiF.34.5 AlF 3 .27.5 PbF 2

Exposure FlowTemperature Time Rate

RAP Gas °C mm Relative cm

None* 9500 - - 1.834

8.04

N2 9500 30 10 .153

N2 9500 45 10 .164

10% NF 3 in N2 9500 15 4 .619

10% NF 3 in N2 9500 30 4 1.694

10% NF 3 in N2 9500 36 4 .534

10% CF 4 in N2 9500 30 10 3.54

10% CCI 2 F2 in N2 9500 15 2 .221

10% CCI 2 F 2 in N2 9500 30 2 .301

10% CC! 2 F 2 in N2 9500 45 2 .09810% CC3 F in N2 9500 15 2 .290

10% CCl 3F in N2 9500 30 2 .094

10% CCIF 3 in N2 950' 15 4 .522

10% CC1F3 in 42 9500 30 4 .177

10% CCIF3 in N2 9500 45 4 .098

Data taken from eleven 30 g melts prepared under dry N2 , andmelted for 15 minutes at 950 0 C.

j

p ~-9-.

I--9

TABLE 4

SPECIFICATIONS AND PROPERTIES OF RAP GASES

Boiling Pt.°C Purity

Gas Supplier (760 Torr) (Grade)

N2 House nitrogen -98%

NF Air Products -129.06 0 C Technical98.0 wt%

CF4 Air Products -128 0 C Semiconductor10% Halocarbon - 14 .n N2 99.9%

CCIF 3 Air Products --81.2 0 C C.P. Grade10% Halocarbon - 13 in N2 99.0% (liquid

phase)

CC1 2 F2 Air Products -29.9 0 C C.P. Grade10% Halocarbon - 12 in N2 99.0% (liquidphase)

4 CF Air Products 23.77 0 C C.P. Grade10% Halocarbon - 11 in N2 99.9% (liquid

phase)

PAT/bb

F- 'RPT001:2

FIGURE 1.

T EM OF QUENCHED 29.5 CdF2 8.5 LiE*34.5 A1F3 27.5 P 2

S ��quipirft

IS

4

- 1.

4 -� 11111,4.

A /

'I2

FIGURE 2.

I VISCOSITY OF 29.5 CdF2 .8.5 LiF 34.5 AIF 3 •12 27.5 PbF2

2860 C

II

SE,= 456kWd/mol

. O00/T (0K)

TRANSMISSION

0 0 0 0 0

Hc

zPco U)

z z

101000

rri0 -z

0

ooe (nrn

r 0,0 0 0r _

0 U) Zr

0

00

.9. rp1i

5

00 0A ~00 ' 0 M

(,g I _ g(° C I.0 -o

00

"0 ----""

0 orn m

C")

z H-I X

04

""1

Ir-n

0 TRNMSSO '

0-a0 m

NJ )

0 0 Iz1"

0 m0

FIGURE 5.CONCENTRATION DEPENDENCE OF ABSORPTION FORSELECTED RARE EARTH ABSORPTION PEAKS IN29.5 CdF2 • 8.5 LiF. 34.5 AIF 3 27 5 PbF 2

Er 1530 nm PEAKSm 1380 nm PEAK

0

I Nd 677nm PEAK

0.-

0.011000 10,000

CONCENTRATION, ppm w

FIGURE 6. VAC GLOVE WX ELEVATOR

MELTING FURNACE OOMR

SCHEMATIC

ANNEALING

ANTE FURNACE

CHAMBER TMPERATURE

OYROMETER

o _

HFATIG~HICLDS

ECCO 0

-3

INDUCTION PLATINUM 0 o

HEATER ~~CRUCIBLE 0RP4I

0 0L

RAIGAEXHAUS'GAS OWl

PURGE %GAS INti_

B.PHOTOGRAPH

FIGURE 7EFFECT OF CCIF3 ON WATER CONTENT 'N 29.5CdF1 "8.5 LiF. 27.5 PbF 2 -34.5 AIF 3

15 MINUTES AT 95-0

030 MINUTES AT 9500

- 4`1 MINUTES AT 9500

80

0 60- 3U i)

- U,U)250 •

z•:40-

20-

10

02 3 4 5 6MICMICRONS

% TRANSMISStON

OD 8N> po

cn C:00

WATER

0 ;u0 >

zro Op

m cnz (n

rn0 0c L Z 0

WATER XOD00 00

ft-wo - CLro T

-n OD

0c;u

]OD