In vitro solubility and in vivo toxicity of gallium arsenide

TOXICOLOGY

AND

APPLIED

PHARMACOLOGY

76,96-104

(1984)

In Vitro Solubility and in Viva Toxicity of Gallium Arsenide’,* D. R. WEBB,~ I. G. SIPES, AND D. E. CARTER Department of Pharmacology and Toxicology, College of Pharmacy, University of Arizona, Tucson, Arizona 85721

Received November 8, 1983; accepted May 8, 1984 In Vitro Solubility and in Vivo Toxicity of Gallium Arsenide. WEBB, D. R., SIPES, I. G., CARTER, D. E. (1984). Toxicol. Appl Pharmacol. 76, 96-104. The in vitro solubilities of gallium arsenide (GaAs) and its metal oxides were arsenic(II1) oxide > GaAs % gallium(II1) oxide. GaAs dissolution was also dependent upon the type. and concentration of buffer anion. The amount of arsenic dissolved in 12 hr by various aqueous media was 0.2 M phosphate buffer 2 0.1 M phosphate buffer z Krebs-Hensleit buffer > distihed Hz0 > HCl-KCl buffer. GaAs was apparently soluble under in vivo conditions. Blood arsenic concentrations in rats 14 days after intratracheal instillation of 10, 30, or 100 mg/kg GaAs were 5.5, 14.3, and 53.6 apl ml, respectively; gallium was not detected at any doses. An increase in lung wet weight at 14 days was dose dependent with these organs retaining 17 to 42% of the dose as gallium or arsenic. Excretion of gallium and arsenic was limited to the feces. Urinary porphyrin concentrations and body weight, monitored as indices of toxicity, were significantly altered over the 14day study. The analysis of porphyrins revealed that uroporphyrin replaced coproporphyrin as the primary urinary metabolite. Rats receiving 10, 100, or 1000 mg/kg GaAs po exhibited similar signs of toxicity. Blood arsenic concentrations at 14 days were 3.5, 6.8, and 17.6 pg/ml, respectively. Porphyria was increased, and body weight was decreased at 1000 mg/kg GaAs. These values were equivalent to those obtained with an intratracheal dose of 10 to 30 mgjkg GaAs. Our results showed that pulmonary and po exposure to GaAs resulted in systemic arsenic intoxication. The finding that urinary uroporphyrin concentrations were greater than coproporphyrin concentrations may serve as a sensitive indicator for GaAs exposure. 0 1984 Academic Ress, Inc. AND

Gallium arsenide (GaAs) is a crystalline, intermetallic compound possessing superior semiconductor properties relative to the more common materials such as silicon or germanium. Although silicon is still the most important material for semiconductor technology, GaAs has led to the development of electro-optical devices such as light-emitting diodes and semiconductor lasers (Kohl and Vieweg-Gutburlet, 198 1). Due to superior

electron flow velocity, GaAs has also been the material of choice for high-frequency microwave and millimeter wave telecommunications systems and ultrafast supercomputers (Robinson, 1983). Exposure to airborne particulates of GaAs is a potential health hazard in the semiconductor industry (Boeniger and Briggs, 1979; Pickrell et al., 1979). However, the degree of risk workers assume with GaAs exposure is not clear since the toxicity of GaAs has not been adequately defined. This study was designed to address this deficiency and to investigate the in vitro dissolution of GaAs as well as the absorption, excretion, and toxicity of GaAs by oral and intratracheal routes of exposure.

’ Presented in part at the Twenty-Second Annual Meeting of the Society of Toxicology, March 11, 1983, Las Vegas, Nev. * Supported by NIEHS Training Grant T32 ES07091 and by the National Toxicology Program (ES821 30). 3 Current address: The Proctor & Gamble Co., 6210 Center Hill Road, Cincinnati, OH 45224. 0041-008X/84

$3.00

Copyright 8 1984 by Academic Prcq Inc. All rights of reproduction in any form reserved.

96

GALLIUM

ARSENIDE

METHODS Preparation of materials. GaAs was purchased from AIf& Products (Thiokol/Ventron Division, Danvets, Mass., purity > 99.99%) and pulverized in a freezer mill (Splex Industries, Inc., Metuchen, N.J.). The powder was sieved through a #400 mesh microsieve in a sonic sifter (AllenBradley, Milwaukee, Wise.). The sievable fraction was analyzed for particle size distribution with a model #I lZCLTNJB/ADC (W) Electrozone/Celloscope (Particle Data, Inc., Elmhurst, Ill.) fitted with a ISO-pm orifice probe. Particle shape was determined by scanning electron microscopy. Arsenic(II1) oxide (AszOr) and gallium(III) oxide (GarOr) were purchased from Alfa Products (Alfa Puratronic, Cat. #lo0106 and #32102, respectively) and similarly sieved and analyzed. The microsieved fraction of all samples was used throughout the study. Analytical determination of metal concentration. Gallium was analyzed by tlame atomic absorption spectrophotometry (FAAS). Absorbance was measured at 294.4 nm during aspiration of samples into an air-acetylene flame. If gallium were not detected by FAAS, samples were analyzed by flame emission spectrophotometry (FES). Emission was measured at 403.7 nm in a nitrous oxide-acetylene flame. For the quantitation of gallium in biological samples, standard curves were prepared in the appropriate tissue or excreta. Arsenic was analyzed by FAAS following formation and separation of the volatile amine (Webb and Carter, 1984). Wet digestion of biological samples. Lungs were homogenized in cold, 0.9% (w/v) saline at a 1:lO (w/v) ratio. A 20% (w/v) fecal homogenate was prepared in 20% (v/v) HNOs. A < 1-ml aliquot of homogenate, urine, or heparinized blood was then transferred to 25 ml Erlenmeyer flasks (4 replicates per sample). Stone chips, approximately 30 mg of K&O,, and 3 ml of concentrated HNOr were added to each flask. Flasks were heated on a hotplate at 120 to 140°C to near dryness. The flasks were removed, cooled, and 3 ml of concentrated H2S04 was added. Gallium samples: Two flasks per sample were returned to the hotplate and heated at 190 to 200°C for 30 min; formation of a precipitate was strictly avoided. Flasks were removed from the hotplate, cooled, and the contents transferred and brought to volume with distilled H,O. If charring occurred during the H2S04 stage, flasks were removed from the hotplate, cooled, and 3 ml of concentrated HNOr was added. Flasks were returned to the hotplate and heated to remove HN09. This step was repeated until charring stopped. Arsenic samples: A more rigorous digestion procedure was required to decompose methylated metabolites of arsenic to its inorganic form. The remaining two flasks per sample were subjected to the wet digestion procedure of Webb and Carter (1984) that has been shown to quantitatively recover total arsenic.

TOXICTI-Y

97

In vitro solubility. A 20 mg sample of GaAs was added to 40 ml of 0.1 M, pH 7.4, phosphate buffer in 125 ml Erlenmeyer flasks. Flasks were placed in a shaking incubator for 48 hr at 37°C. Aliquots were removed at various time points from 0.25 to 48 hr and filtered through 0.2-pm filters (Gelman Acrodisc, Gelman, Ann Arbor, Mich.). The samples were then diluted with 6 N HCl and 4% (w/v) Nal (2: 1:1). Equimolar amounts, on a metals basis, of Asr03 and Gar03 were similarly incubated, sampled, and diluted. Filtrates were analyzed for arsenic and/or gallium content by FAAS. Arsenic concentrations were quantified against an arsenic(lI1) standard. Gallium concentrations were determined from a standard curve prepared in a similar aqueous matrix. GaAs was also incubated for 12 hr under identical conditions in the following aqueous systems:0.2 M phosphate buffer, pH 7.4; 0.1 M phosphate buffer, pH 7.4; Krebs-Hensleit buffer, pH 7.4 (Moldeus et al., 1978); HCl-KC1 buffer, pH 2.0; and distilled HrO. Samples were diluted, and arsenic concentrations were determined in the filtrates as above. Treatment of animals. Male Fischer-344 rats (190 to 2 10 g) were obtained from Charles River Laboratories, Boston, Mass., and acclimated for 7 days prior to dosing. Rats were allotted to eight groups (n = 3) such that the group means and standard deviations in body weight were approximately equal. Under ether anaesthesia, rats received 10, 100, or 1000 mg/kg GaAs po or 10, 30, or 100 mg/kg GaAs by intratracheal instillation. Control rats received a sham dose of vehicle by the appropriate routes of administration. GaAs was suspended in Iihersterilized normal saline and delivered within 5 min of saline addition in a 1.0 ml volume po or 0.7 ml intratracheally. This procedure ensured that a minimal amount of the GaAs dose would be in a soluble form at the time of administration. Suspensions were delivered intratracheally with a 20-gauge X 1.5 in. needle after surgical exposure of the trachea. Approximately 0.35 ml was directed toward each bronchus and the incision closed. Rats were housed in individual metabolism cages on Days 1 to 7 and on Day 14. Urine was collected over 1 ml of 1.0% (w/v) NaF-2.0% (w/v) NaHCOr to reduce bacterial growth and to oxidize porphyrinogens to porphyrins. Feces were also collected and excreta stored at -1O’C. Body weight was recorded daily. On interim Days 8 to 13, rats were group housed in shoe-box cages. Water and feed (Wayne Lab Blox) were provided ad libitum throughout the study. A lo-14 hr light-dark cycle was maintained. At the end of the 14&y study, rats were anaesthetized with sodium pentobarbital, ip. Blood was removed from the superior vena cava, heparinized, and stored at - 10°C. Major organs were removed, rinsed in cold normal saline, blotted dry, and weighed. Organs were examined for gross pathological lesions. The liver and kidneys of animals receiving 100 mg/kg GaAs intratracheally were also examined histopathologically using H St E stain.

98

WEBB, SIPES, AND CARTER

Analysis of urinary porphyrins. Urine was adjusted to pH 8 to 9, if necessary, with NaHCDs and incubated at room temperature for 8 to 12 hr. The urine was diluted with 29 volumes of a solution of 1.5 M citric acid containing 0.1 M oxalic acid. Samples were mixed and allowed to stand in the dark at room temperature for 20 min. The diluted urine was filtered through a 0.45-pm Millex-HA filter (Millipore Corp., Bedford, Mass.) prior to analysis. Porphyrins were separated on a C,s Bondapak column (Waters Associates, Milford, Mass.) with a gradient solvent system of pH 3.5,O. 1 M NH.,H$Q:MeDH (55:45, Bottle A) and MeGH (Bottle B). Fluorescence was measured at I0 398 nm and IP 625 nm, and values quantified against a standard porphyrin mixture (Porphyrin Products, Logan, Ut.). Treated groups served as their own controls during the 24-hr period prior to dosing. Statistical analyses. A statistically sign&ant difference (p < 0.05) between groups was established by ANOVA and least significant difference tests. Pooled porphyrin values were analyzed relative to a 95% confidence interval calculated from nonpooled control values.

LOQ kW

(houTS)

1. The concentration of arsenic and/or gallium in 0.1 M, pH 7.4, phosphate buffer liltrates during the incubation of GaAs, A%Os, or GarOs at 37°C for 48 hr. Each value represents X + SD of 3 replicates. FIG.

RESULTS Particle size data. The mean count diameter for GaAs, Ga203, and As203 particulates was 8.30, 6.80, and 9.17 pm, respectively. Particle shape was roughly spherical for all samples. The mean volume diameter, in the same order of presentation, was 12.67, 16.43, and 18.58 pm. In vitro solubility. GaAs was apparently soluble in 0.1 M, pH 7.4, phosphate buffer. The results are shown in Fig. 1. Arsenic concentration in buffer filtrates peaked at 200 &ml by 36 hr which represented 78% dissolution. These values were equal to those obtained with AszOs suggesting that the buffers became saturated with arsenic. However, the amount of As203 dissolution was substantially greater than that observed with GaAs during the first hour of incubation. The linear release of arsenic was analyzed by a power function equation: y = mxb where y is pg arsenic/ml, m is the intercept value at 1 hr, x is the time in hr, and b is the slope. The intercept and slope values for arsenic release from GaAs during 0.25 to 36 hr incubation were identified as y = 7.82x0.*9 with a, correlation coefficient of 1.OO. Similar analysis for arsenic released from Asz03 iden-

tified these values as y = 37.68~‘-~’ with a correlation coefficient of 0.97. The concentration of gallium released horn GaAs was maximal at 150 &ml by 48 hr which represented 62% dissolution (Fig. 1). However, gallium values were still increasing at this time. Power curve analysis results were y = 4.88~“.*~ with a correlation coefficient of 0.97. Gaz03 was very insoluble as < 1% dissolution had occurred by 48 hr (Fig. 1). The in vitro dissolution of GaAs was affected by the composition of the aqueous incubation medium. The amount of arsenic released in 12 hr appeared to be dependent upon the amount and type of anion present. When arsenic concentration was normalized to the amount released from 0.1 M phosphate buffer, we found that 0.1 M phosphate buffer, 100 + 20% (2 f SD); 0.2 M phosphate buffer, 122 + 11%; Krebs-Hensleit buffer, 56 + 3%; distilled water, 40 + 5%; and HCl KC1 buffer, 11 + 2%. Absorption, excretion, and pulmonary retention of GaAs. Absorption of GaAs was investigated by measuring blood concentrations of gallium and arsenic on Day 14.

GALLIUM

ARSENIDE TABLE

99

TOXICITY

1

RECOVERYOF GALLIUM AND ARSENIC IN SELECTED TISSUESAND EXCRETA: INTRATRACHEAL STUDYa Dose recovered (%) Gallium arsenide h3/W loo 30 10

Urine

FeCeS

Blood

Lungs

Arsenic Gallium

0.13 zk 0.02 ND”

12.8 19.1

10.4 * 2.1

Arsenic Gallium

ND ND

19.9 33.0

9.2 + 1.4 ND

16.5 +- 4.4 22.7 -e 5.3

Arsenic Gallium

ND ND

9.3

10.6 f 1.7

ND

ND

17.5 +- 4.8 25.4 +- 6.2

32.5 +- 9.5

ND

41.8 +- 12.5

a Rats received intratmcheal instillations of gallium arsenide suspended in normal saline. Excreta were collected during Days 1 to 7 and 14. Lungs and blood samples were taken on Day 14. Each value represents X k SEM of 3 tats except for fecal data which represents a pooled sample. b Not detected (a detection limit of 10 ppm in urine and 20 ppm in blood was obtained).

Arsenic blood values (2 + SEM) for rats dosed intratracheally with 10, 30, or 100 mg/kg GaAs were 5.5 f 0.9, 14.3 f 2.1, and 53.6 -+ 10.9 &ml, respectively. These values rep resent 9 to 10% of the total arsenic in GaAs (Table 1). Rats receiving 10, 100, or 1000 mg/kg GaAs po had blood arsenic concentrations of 3.5 + 0.4, 6.8 + 0.5, and 17.6 +- 2.8 &ml, respectively. These values represent a range of 0.3 to 7.0% of the dose. Gallium was not detected in blood at any dosage by either route of administration (Table 2). Excretion of gallium and arsenic was primarily confined to the feces regardless of the route of GaAs administration (Tables 1 and 2). Trace amounts of arsenic were observed in the urine from rats receiving 1000 m&kg GaAs po or 100 mg/kg GaAs intratracheally. Gallium was not detected in urine at any dosage by either route of administration. Fecal recovery of arsenic ranged from 9 to 20% of the dose following intratracheal delivery of GaAs and 56 to 9 I YO of the dose PO. The recovery of gallium in feces ranged from 19 to 33% and 70 to 99% of the dose, respectively. Fecal elimination of both metals was complete by Days 3 to 4. GaAs was retained in the lungs of rats 14 days after the instillation of 10, 30, or 100

mg/kg (Table 1). Arsenic retention ranged from 17 to 32% of the dose while gallium retention ranged from 23 to 42% of the dose. Indices oftoxicity. The intratracheal instillation of GaAs affected body weight over the 14&y study (Fig. 2). Body weight was lost during the first 24 hr following dosing. Control animals quickly recovered from surgery and saline instillation to subsequently gain weight at a steady rate. However, GaAs instillation reduced the amount of weight gain normally expected during Days 1 to 7 with TABLE 2 RECOVERYOF GALLIUM AND ARSENICIN BLOOD AND EXCRETA: ORAL STUDY a Dosereeovcred(W)

Gallium arsenide bwh8)

Iwo 100 IO

Urine

Arsenic Gallium

0.02 + 0.01

Arsenic

ND

Gallium

ND

AlSCOiC

ND

Gallium

ND

NDb

FeceS

BlOCId

90.7 + 35.4 99.4 + 38.7

0.3 k 0.02

79.8 f 89.2 f

4.2 5.4

1.3 i 0.1 ND

56.0 f 70.3 k

2.8 6.0

7.0 + 1.1

ND

ND

E Rats received suspensions of gallium arsenide in normal saline. Excreta were collected during Days 1 to 7, and 14. Blood samples were collected on Day 14. Each value represents the X + SEM of 3 rats. bNot detected (a detection limit of 10 ppm in tuioe and 20 ppm in blood was obtaioed).

100

WEBB, SIPES, AND CARTER 2

32

1

8

-16

4 0

2

4

6

I 8

(I 14

Time (days)

FIG. 2. The cumulative 14day body weight change in rats following the intratracheal instillation of normal saline (control) or GaAs suspensions. Each value represents f f SD of 3 rats. significantly different from control (p i 0.05).

the depression being maximal and significant (p < 0.05) at 100 mg/kg. During Days 7 to 14, treated rats gained weight at a rate equal to control animals although they were unable to regain the body weight initially lost in the first 7 days of the study. Rats receiving GaAs po displayed a similar pattern in body weight change. Body weight data for rats dosed with 10, 100, or 1000 mg/kg GaAs po (data not shown) were equivalent to those observed in animals intratracheally instilled with 10 or 30 mg/kg and were not significantly different from control (p > 0.05) at any time point throughout the study. Major organs examined on Day 14 for wet weight and pathological lesions were the heart, lungs, liver, spleen, kidney, and testes. With the exception of lungs, all organs, regardless of dose or route of administration, were pathologically unremarkable and similar in wet weight to those of control animals. Lungs from rats intratracheally dosed with GaAs were dramatically affected. Retention of GaAs appeared as grey, focal areas throughout the lung. When the wet weight of these organs was expressed as a lung weight:body weight ratio, GaAs caused a significant (p < 0.05) and dose-dependent increase relative to control (Fig. 3). It should be noted that the amount of the dose remaining represented

an insignificant proportion of the total lung weight change. Lungs obtained from rats that received GaAs orally were not similarly a& fected. The intratracheal instillation of GaAs caused a significant (p -C 0.05) increase in the urinary excretion of porphyrins (Fig. 4). The increase was maximally affected at the 100 mg/kg intratracheal dose with values peaking on Day 6. At this dose and time, the increase in urinary porphyrins was 22 1% above the control value of 7.59 f 0.95 &lo0 ml urine (X f SEM, n = 12). A similar but non-dosedependent response in porphyria was observed at 10 and 30 mg/kg GaAs. Rats receiving 1000 mg/kg GaAs po also developed porphyria. In this case, the increase in porphyrin excretion was biphasic. An early peak appeared on Day 1 and an additional peak appeared on Day 6. Values for Days 2 to 14 were similar to those from rats receiving 30 mg/kg intratracheally. Urine samples from 1.0

0.8

0.6

0.4

0.2

0

3 Dose (mg/kg)

FIG. 3. The lung wet weight:body weight ratios (g/100 g) on Day 14 of rats intratracheally instilled with saline (control) or GaAs suspensions. Each value represents zi k SD of 3 rats. Significantly different from control (p < 0.05).

GALLIUM

ARSENIDE

Tma (days)

4. The concentration of porphyrins in urine over 14 days following the intratracheal and oral instillation of GaAs. Each value represents a pooled sample from 3 rats with the exception of the control value which represents 2 + SEM of 12 rats 24 hr prior to dosing (intratracheal, i.t.). FIG.

rats receiving 10 or 100 mg/kg GaAs po were not analyzed for porphyrin. The increased concentrations of urinary porphyrins were predominantly due to the increase in uroporphyrin concentrations (Table 3). The intratracheal instillation of 100 mg/kg GaAs caused uroporphyrin concentrations to increase on Day 6 by a factor of 3.8 while coproporphyrin concentrations increased 1.9 times control values. Similar but lesser increases in uroporphyrin concentrations were also observed at 10 and 30 mg/kg GaAs while coproporphyrin concentrations remained relatively unchanged. Following the 1,000 mg/kg po dose of GaAs, uroporphyrin concentrations were elevated 2.6 times the control values while coproporphyrin concentrations were not significantly affected. The total volume of urine excreted during the 24hr collection periods was approximately equal at all doses. DISCUSSION Studies have identified GaAs as a potential occupational hazard of unknown toxicity following pulmonary exposure to the airborne

101

TOXICITY

dusts (Boeniger and Briggs, 1979; Pickrell et al., 1979). Although these reports have focusd upon its use in the photovoltaic industry, GaAs dust generation is not unique to the manufacture of solar cells. GaAs particles can be generated during procedures involved with the growth and processing of GaAs ingots such as ingot sandblasting and cleaning, ingot cropping, wafer slicing, and wafer polishing (Briggs and Owens, 1980). Therefore, the potential for occupational exposure to GaAs dust exists whenever GaAs production and processing occurs. The aqueous solubility of GaAs was of primary interest in this study. Dissolution of GaAs would release gallium and arsenic species that may have significant local and systemic toxic effects. The aqueous solubility of GaAs particulates has not been reported. GaAs exists in crystalline form with a zincblend lattice and predominantly covalent bonds (Folberth, 1962). Such physicochemical properties may be expected to significantly reduce, if not prevent, aqueous dissolution. This hypothesis was supported by Pickrell et TABLE 3 EFFETE OF GALLIUM PORPH~UN

ARSENIDE ON URINARY LEVELS IN RATS’

Urinary porphyrin concentrations on Day 6 (percentage of control) bf~

be/kg) 10” 30b 100b 1OOOd

Uroporphyrin 336’ 307’ 377’ 259’

Coproporphyrin 115 112 193’ 131

’ Rats received normal saline suspensions of gallium arsenide. Animals were placed in individual metabolism cages and urine collected every 24 hr over 1.0 ml of 1.0% (w/v) NaF-2.0% (w/v) HaHCOs. Individual urine samples were pooled for porphyrin analysis. Actual uroporphyrin and coproporphyrin concentrations (mean t- SEM, n = 12) in control rats were 1.60 f 0.15 and 3.78 +- 0.52 l(g/lOO ml urine, respectively. b Intratracheal instillation. c Significantly different from control (p < 0.05). d Oral administration.

102

WEBB,

SIPES,

al. (1979) who reported that bulk, crystalline GaAs was quite inert in water. However, GaAs forms oxides with exposure to air (Rosenberg, 1960; Aitken, 1967) while the presence of nascent hydrogen in water also favors formation of hydrates, hydroxides, and gaseous hydrides (Aitken, 1967). It is also known that the dissolution rate of many compounds is inversely proportional to particle size. Dissolution is indicated by our in vitro results. Aqueous incubation of GaAs particles (3 volume diameter 12.67 pm) rapidly released gallium and arsenic species capable of passing a 0.2-pm filter. The mechanism and chemistry of this dissolution are unknown but could have resulted from particulate GaAs less than 0.2 pm in size, dissolution of GaAs molecules, or formation of the oxides, GaZ03 and AszO,, originating from an oxidation of A$-III) to As(III) followed by hydrolysis. The Ga:As ratio of the dissolved species was not unity as would be expected if GaAs particles had remained intact or if GaAs molecules had dissolved. The fact that more arsenic was recovered in solution may have resulted from Ga forming an insoluble precipitate at pH 7.4 (Dudley and Levine, 1949) and some of the particles being unable to pass a 0.2-pm filter. We have also shown that the amount of arsenic released in 12 hr appeared to depend on the composition of the medium: phosphate concentration, pH, and chloride ion. These results suggest that the dissolution of GaAs is a complex process that may involve the redox potential in the aqueous medium as an important factor. Dissolution of GaAs following intratracheal and po administration is also indicated by our in vivo results. Significant amounts of arsenic were seen in the blood without detection of any gallium. For example, blood arsenic concentrations reached 54 ppm 14 days after 100 mg/kg GaAs was given by intratracheal administration. Gallium was not detected in the blood during this experiment even though 50 ppm gallium was significantly greater than our detection limit. At the same

AND

CARTER

time, more gallium than arsenic was found remaining in the lungs and excreted in the feces of these rats. With the release of gallium and arsenic from GaAs, one might expect development of their respective signs of systemic toxicity. Gallium is relatively nontoxic when administered by noninvasive techniques. Previous work has shown that soluble salts (e.g., gallium chloride, gallium lactate, gallium citrate) were not readily absorbed by rats following oral (Dudley and Levine, 1949; Dudley et al., 1950) or inhalation exposure (Dudley and Levine, 1949). In contrast, gallium salts administered iv or ip resulted in systemic exposure which produced acute renal toxicity (Dudley et al., 1950; Newman et al., 1979). Our results agree with these studies. We did not detect any gallium in blood or urine. Histologic examination of the kidneys at 14 days following intratracheal instillation of 100 mg/kg GaAs did not show any of the signs of renal toxicity reported by Newman et al. (1979) following ip administration of gallium nitrate. The gallium found in the lung and feces in our study probably resulted from the retention of GaAs particulates and insoluble hydrated gallium oxides that were not absorbed. In contrast to gallium, arsenic was absorbed after po or intratracheal administration of GaAs. Although blood concentrations were as high as 54 ppm, only trace amounts of this element were recovered in urine over the 14day study. The concentration of arsenic in rat blood 14 days after GaAs administration can be viewed as a cumulative value since the rat is recognized as an atypical mammalian model for arsenic toxicity (NAS, 1977). Specifically, arsenic strongly binds to the erythrocytes in the rat but not in other species studied (Hunter et al., 1942; Odanaka et al., 1980; Vahter, 1981). In conjunction with arsenic sequestering with the erythrocyte, urinary excretion of arsenic is signi6cantly lower in the rat than in other species studied (Odanaka et al., 1980; Vahter, 1981). Substantially less GaAs was absorbed after

GALLIUM

ARSENIDE

po administration than after intratracheal administration. Blood concentrations were lower and fecal recoveries were higher when equivalent doses were given orally. For example, the arsenic blood concentration was 1.3% of the 100 mg/kg po dose after 14 days while it was 10.4% of the same dose when given intratracheally. This difference was probably a result of reduced dissolution in the gut as soluble arsenic compounds would be readily absorbed from the gastrointestinal tract if they were formed (Stevens et al., 1977; Odanaka et al., 1980; Vahter, 1981; Rowland and Davies, 1982). Fecal recovery of arsenic from intratracheally dosed rats may have been partially due to some ingestion of GaAs during dosing and pulmonary clearance of the particles followed by po ingestion (Kilbum, 1977). Fecal excretion of systemic arsenic would be small because of extensive enterohepatic recirculation of arsenic excreted in the bile (Klaassen, 1974). Our results showed an increase of porphyrin excretion and an enhanced amount of uroporphyrin over coproporphyrin. This increase has been found for both trivalent and pentavalent arsenic in rats (Woods and Fowler, 1978; Martinez et al., 1983). These results are of interest because the onset of this particular type of hepatic porphyria (Tschudy, 1974) occurred at relatively low concentrations of arsenic exposure prior to the onset of frank hepatotoxicity. We were also unable to detect any signs of hepatic toxicity by histopathologic methods at the doses administered. These data suggest that the specific elevation of uroporphyrin excretion may serve as an important, early indicator for GaAs exposure. Finally, our data demonstrated that the intratracheal instillation of GaAs was more toxic than the po route of administration. A similar observation was reported by Roschina ( 1966). We have shown that a maximal effect in porphyria, body weight depression, and blood arsenic concentrations was observed with 100 mg/kg GaAs when delivered intratracheally. However, it should be emphasized

103

TOXICITY

that these biological parameters appeared to correlate with blood arsenic concentrations regardless of the route of administration of GaAs. A dosedependent increase in lung wet weight 14 days after intratracheal instillation of GaAs was also observed. This response may be associated with the retention of GaAs particulates rather than the GaAs solubility products previously discussed. It is of interest to note that silica has been associated with an increase in lung wet weight that leads to the subsequent development of lung fibrosis (Chvapil et al., 1979; Dauber et al., 1980). Lung fibrosis has also been attributed to the pulmonary exposure of GaAs particulates (Tarasenko and Fadeev, 1980) although this effect was not rigorously documented. ACKNOWLEDGMENTS The authors acknowledge the generous contributions of Beth Gottung and Dr. Alan Randolph (Dept. of Chemical Engineering, University of Arizona) for the analysis of particle size distribution and Dr. Samuel Schwartz (Minneapolis Medical Research Foundation, Inc., Minneapolis, MN) for the analysis of urinary porphyrin concentrations.

REFERENCES AITKEN, E. A. (1967). Corrosion behavior. In Intermetallic Compounds (J. H. Westbrook, ed.), pp. 491-493. Wiley, New York. BOENIGER, M., AND BRIGGS, T. (1979). Potential health hazards in the manufacture of photovoltaic solar cells. In Health Applications of New Energy Technologies (W. N. Rom and V. E. Archer, eds.), pp. 593-606. Ann Arbor Science Publishers, Ann Arbor. BRIGGS, T. M., AND OWENS, T. W. (1980). Industrial Hygiene Characterization of the Photovoltaic Solar Cell Industry. NIOSH Technical Report, DHEW (NIOSH) Publication No. 80-112, U.S. Department of Health, Education, and Welfare, Cincinnati, OH 45226. CHVAPIL, M., E!SKEL.SON, C. D., STIFFEL,V., ANDOWENS, J. A. (1979). Early changes in the chemical composition of the rat lung after silica administration. Arch. Environ. Health 34,402-406. DAUBER, J. H., ROSSMAN, M. D., PIETRA, G. G., JIMENEZ, S. A., AND DANIELE, R. P. (1980). Experimental silicosis; morphologic and biochemical abnor-

104

WEBB, SIPES, AND CARTER

mahties produced by intmtracheal instigation of quartz into guinea pig lungs. Amer. J Pathol. 101, 595-607. DUDLEY, H. C., AND LEVINE, M. D. (1949). Studies of the toxic action of gallium. J. Pharmacol. Exp. Ther. 95,487-493. DUDLEY, H. C., HENRY, K. E., AND LINDLSEY, B. F. (1950). Studies of the toxic action of gallium. II. J. Pharmacol. Exp. Ther. 98, 409-417. FOLBERTH, 0. G. (1962). The chemical bond in the IIIV compounds. In Compound Semiconductors (R. K. Willardson and H. L. Georing, eds.), pp. 21-33. Reinhold, New York. HUNTER, F. T., KIP, A. F., AND IRVINE, J. W., JR. (1942). Radioactive tracer studies on arsenic injected as potassium amen&e. J. Pharmacol. Exp. Ther. 76, 207-220. K~LBURN, K. H. (1977). Clearance mechanisms in the respiratory tract. In Handbook in Physiology, Section 9 (D. H. K. Lee, H. L. Falk, S. D. Murphy, and S. R. Geiger, eds.), pp. 243-262. Bethesda. KLAASSEN, C. D. (1974). Bihary excretion of arsenic in rats, rabbits, and dogs. Toxicol. Appl. Pharmacol. 29, 447-457. KOHL, F., AND VIEWEG-GUTE~ERLET,F. G. (198 1). Gallium arsenide-the material and its application. Microelec. J. 12, 58. MARTINEZ, G., CEBRIAN, M., CHAMORRO, G., AND JAUGE, P. (1983). Urinary porphyrins as an indicator of arsenic exposure in rats. Proc. West. Pharmacol. sot. 26, 171-174. MOLDEUS, P., HOGBERG, J., AND ORRENIUS, S. (1978). Isolation and use of liver cells. In Methods in Enzymology (S. Fleischer and L. Packer, eds.), pp. 60-71. Academic Press, New York. NAS (1977). Medical and Biological Effects of Environmental Pollutants: Arsenic. National Academy of Sciences, Washington, D.C. NEWMAN, R. A., BRODY, A. R., AND KRAKOPF, I. H. (1979). Galhum nitrate (NSC-15200) induced toxicity in the rat. Cancer 44, 1728-1740. ODANAKA, Y., MATANO, O., AND GATO, S. (1980). Biomethylation of inorganic arsenic by the rat and some laboratory animals. Bull. Environ. Contam. Toxicol. 24, 452-459.

P~CKRELL, J. A., HOBBS, C. H., AND MOKLER, B. V. (1979). Biomedical research needs related to photovoltaic conversion of solar energy to electricity. Inhalation Toxicology Research Institute, Lovelace Biomedical and Environmental Research Institute, Albuquerque, NM. ROBINSON, A. L. (1983). GaAs readied for high-speed microcircuits. Science 219,275-277. ROSCHINA,T. A. ( 1966). ToxicologicaI features of indium and gallium arsenide-a new group of semiconductors. Gig. Tr. Proj Zabol. 10, ISS 5, 30-33. ROSENBERG,A. J. ( 1960). The oxidation of intermetallic compounds. J. Phys. Chem. Solids 14, 175-180. ROWLAND, I. R., AND DAVIES, M. J. (1982). Reduction and methylation of sodium arsenate in the rat. J. Appl. Toxicol. 2, 294-299. STEVENS,J. T., HALL, L. L., FARMER, J. D., DIPASQUALE, L. C., CHERNOFF, N., AND DURHAM, W. F. (1977). Disposition of “C and/or “As-cacodylic acid in rats after intravenous, intmtracheal, or peroral administration. Environ. Health Perspect. 19, 151-157. TARASENKO, N. Y., ANDFADEEV, A. I. (1980). Problems of industrial hygiene related to the use of gallium and indium compounds in the national economy. Gig. Sanit ISS 10. 13-16. TSCHUDY, D. P. (1974). Porphyrin metabolism and the porphyrias. In Duncan’s Diseases of Metabolism, (P. K. Bondy and L. E. Rosenberg, eds.), 7th Ed., pp. 775-824. W. B. Saunders, Philadelphia. VAHTER, M. (198 1). Biotransformation of trivalent and pentavalent inorganic amenic in mice and rats. Environ. Res. 25, 286-293. WEBB, D. R., AND CARTER, D. E. (1984). An improved wet digestion procedure for the analysis of total arsenic in biological samples by direct hydride flame atomic absorption spectrophotometry. J. Anal. Toxicol. 8, 118-123. WOODS, J. S., AND FOWLER, B. A. (1978). Altered regulation of mammalian hepatic heme biosynthesis and urinary porphyrin excretion during prolonged exposure to sodium arsenate. Toxicol. Appl. Pharmacot. 43, 361-371.